Association between women’s empowerment and mental health help-seeking behaviour in Bangladesh: findings from a nationally representative survey

Abstract

Background

Despite a high disease burden, care-seeking for mental health symptoms is low in Bangladesh, particularly among women.

Objective

To evaluate help-seeking behaviours for anxiety and depression symptoms among Bangladeshi women aged 15–49 and its association with women’s empowerment.

Methods

Data from the Bangladesh Demographic and Health Survey 2022 including 2881 women reporting anxiety or depression symptoms were analysed. Help-seeking behaviours, sociodemographics and women’s empowerment using the Survey-based Women’s Empowerment (SWPER) index were assessed. Logistic regression models were used to explore associations.

Results

The mean age of participants was 33.3 years (SD 8.7); two-thirds lived in rural areas. In the SWPER index, 83% of women had high empowerment in the attitude to violence domain, 16% in social independence and 60% in decision-making. Among women with anxiety or depression symptoms, 18.7% sought help, mainly from spouses (43%) or family (52%), and 15% consulted doctors. Women aged >30 years (adjusted OR, aOR 1.25, 95% CI 1.02 to 1.53), from rich households (aOR 1.49, 95% CI 1.09 to 1.89), and with secondary (aOR 1.47, 95% CI 1.08 to 2.01) or higher education (aOR 1.56, 95% CI 1.03 to 2.36) were more likely to seek help. High empowerment in attitude to violence (aOR 0.54, 95% CI 0.37 to 0.81) and decision-making (aOR 0.53, 95% CI 0.41 to 0.68) correlated negatively with help-seeking.

Conclusions

Few Bangladeshi women seek help for mental health symptoms, relying on informal sources. Age, education, wealth and employment predict help-seeking, while high empowerment level in attitude to violence and decision-making domain of SWPER index shows an inverse relationship.

STRENGTHS AND LIMITATIONS OF THIS STUDY

• This study used a nationally representative dataset from Bangladesh Demographic and Health Survey, 2022.

• Mental health symptoms were assessed using validated screening tools, such as Patient Health Questionnaire-9 for depressive symptoms and Generalised Anxiety Disorder-7 for anxiety symptoms.

• The SWPER index was used for assessment of women empowerment, a globally validated tool, although it does not accommodate broader dimensions of women’s empowerment.

• Help-seeking behaviour was self-reported and may be subject to recall or social desirability.

Background

Mental health is increasingly recognised as a global priority.¹ In the Sustainable Development Goals, it has been explicitly highlighted as an essential component of well-being and a comprehensive action was warranted to address this critical issue.² Globally, mental disorders affect nearly 970 million people, contributing almost 5% to total disability-adjusted life-years, with anxiety and depression as the predominant contributors.³

In Bangladesh, mental health disorders pose a significant public health challenge. It is estimated that around one in five adults in the country experiences a mental health condition, with anxiety and depression being the most common.^{4 5} Women are more vulnerable to experience mental health issues with a higher prevalence of 21.5% compared with 15.7% in men.⁴ Several biological, social and economic factors contribute to this disproportionate risk of mental health disorder among women.⁶ In a lower-middle-income country like Bangladesh, gender-based discrimination and violence, limited access to educational and employment opportunities, and societal pressures and expectations regarding marriage and family roles further compound mental health risks for women.6,8

Despite the high prevalence of mental disorders in Bangladesh, a significant unmet need for mental healthcare exists, with particularly limited access for women.⁵ The National Mental Health Survey of Bangladesh (2019) reported that more than 90% of women who were screened positive for mental health symptoms did not seek help during their lifetime; and among the few who sought help, the majority relied on informal sources, such as family members, friends and neighbours, rather than formal healthcare providers.⁴ In addition, Bangladesh encounters multiple structural barriers to mental healthcare. These include a shortage of mental health professionals, such as psychiatrists, psychologists, mental health nurses and other allied specialists, along with high treatment costs and limited infrastructure. Together, these factors severely limit access to mental health services.⁵ Additionally, stigma surrounding mental health, fear of social discrimination and lack of awareness regarding mental health symptoms contribute to a culture of silence and reluctance to seek professional help.^{4 9} In addition to structural barriers, existing stigmatised societal norms and beliefs surrounding mental disorders, restrictive gender roles, financial dependency and limited autonomy further marginalise women, impeding their access to mental healthcare.10,12 Regional evidence from south-east Asia also reported that stigma, low mental health literacy, cultural beliefs and poor distribution of resources are dominant barriers to accessing mental health services across countries in the region. These challenges are often compounded for women and individuals from rural or marginalised communities.^{13 14} In this context, empowerment of women can influence their healthcare-seeking behaviours, as empowered women are more likely to seek help for health issues and advocate for their own needs.¹⁵ A prior study from Nepal suggested that women’s empowerment, particularly in the domain of decision-making autonomy, is positively associated with their mental healthcare-seeking behaviour.¹⁶

In Bangladesh, while past studies have shown a positive link between women’s empowerment and access to maternal healthcare,^{17 18} the relationship between empowerment and seeking mental healthcare might be more complex. Factors like broader social and cultural influences play a significant role. Although empowerment can enhance women’s autonomy and promote positive attitudes towards mental healthcare, the stigma surrounding mental illness remains a major obstacle. Given the multiple factors influencing mental health care-seeking behaviours and the complex role of empowerment, further research is necessary to better understand how women’s empowerment affects their willingness to seek help for mental health concerns.

Objective

The objective of the present study was to assess the help-seeking behaviour for symptoms of anxiety or depression among women aged 15–49 years in Bangladesh and its association with women empowerment status from a nationally representative sample.

Methods

Data source and participants

In the present study, we used the data from Bangladesh Demographic and Health Survey (BDHS), 2022.¹⁹ The survey employed a two-stage, stratified sampling procedure to select enumeration areas (EAs) across urban and rural areas from eight administrative divisions of the country. In the first stage, a total of 675 EAs were selected, 237 from urban areas and 438 from rural areas, using a probability proportional to size approach. In the second stage, 45 households were systematically selected from each EA. All ever-married women aged 15–49 years within these households were identified as eligible respondents for the individual interview portion of the survey. Among the 45 households, 30 were randomly chosen to participate in a detailed questionnaire, while 15 received a shorter version of the questionnaire. The detailed questionnaire included a mental health module, which incorporated standardised screening tools to assess symptoms of anxiety and depression, as well as questions on help-seeking behaviours and treatment history. Although BDHS methods did not report a separate response rate for the mental health module, it was part of the detailed questionnaire, for which they achieved a high overall response rate of 98.9%, with 98.3% in urban areas and 99.2% in rural areas. However, the number of contact attempts per respondent to minimise the non-response rate was not explicitly mentioned in the methodology.

The survey used Generalised Anxiety Disorder-7 (GAD-7) and Patient Health Questionnaire-9 (PHQ-9) scale to assess symptoms of anxiety and depression respectively. GAD-7 is a reliable seven-item tool for identifying persistent worry and symptoms of related anxiety disorders such as panic disorder, social anxiety and post-traumatic stress disorder.²⁰ On the other hand, PHQ-9 is a nine-item tool for assessing the severity of depression symptoms based on Diagnostic and Statistical Manual of Mental Disorders criteria. Each question corresponds to a core symptom of major depressive disorder, such as loss of interest, changes in sleep and feelings of worthlessness.²¹ In the BDHS survey, respondents were classified as having symptoms of anxiety with a GAD-7 score of 6 or higher and symptoms of depression with a PHQ-9 score of 10 or higher. We also used this cut-off as the inclusion criteria in our study.

Our primary sampling frame included a total of 16 046 women with complete data on the PHQ-9 and GAD-7 screening tools. Among these, 4.4% screened positive for depressive symptoms and 17.2% screened positive for anxiety symptoms. A total of 2881 women who reported a GAD-7 score of 6 or higher and/or a PHQ-9 score of 10 or higher during the preceding 2 weeks of the survey were included in our final analysis. We excluded participants with missing data in the mental health modules from our analysis.

Patient and public involvement

Patients and/or the public were not involved in the design, or conduct, or reporting, or dissemination plans of this research.

Variables

Dependent variable

In this study, we considered the help-seeking behaviour of women experiencing symptoms of anxiety or depression as the dependent variable. The help-seeking behaviour was assessed through the question: ‘Thinking about what you yourself have experienced among the different things we have been talking about, have you ever tried to seek help?’ Respondents who answered ‘yes’ were categorised as having sought help, whereas those who answered ‘no’ were categorised as not seeking help. The intended time frame of help seeking was 2 weeks, consistent with the reference period of symptom assessment in the PHQ-9 and GAD-7 instruments. However, this time frame was not explicitly stated in the BDHS 2022 documentation.

Women who reported seeking help were subsequently asked to specify from whom they sought the help. Additionally, information on whether respondents had taken any medication for their symptoms in the preceding 2 weeks was also collected. For those taking medication, further inquiry determined if it had been prescribed by a doctor or other qualified health professionals (defined as professionals who provide medical care at public or private hospitals).

Independent variables

We included independent variables such as age of the respondents, area of residence (urban or rural), wealth quintile, educational attainment, employment status and women empowerment level. Wealth quintile was measured from a composite index based on household assets, living conditions and utility access, segmented into five levels from poorest to richest. Educational attainment was categorised into four groups: no formal education, primary, secondary and higher secondary or above. Employment status was categorised by identifying whether respondents were employed during a period of 12 months preceding the survey or not.

To assess the women empowerment level, we used the SWPER index.^{22 23} This index evaluates the level of women empowerment across three domains: attitude to violence, social independence and decision-making. The ‘attitude to violence’ domain covered women’s perspectives on the justification of domestic violence, whereas ‘social independence’ included aspects like education, media engagement, age at first childbirth and cohabitation, and age and education gaps with partners. ‘Decision-making’ focused on participation in household decisions, highlighting autonomy within the home. The Survey-based Women’s Empowerment (SWPER) index score was derived through principal component analysis (PCA) of 14 items across 62 Demographic and Health Surveys (DHS) surveys. For women without childbirth histories, the age of first birth was estimated via single hot-deck imputation. Using PCA factor loadings, each item was weighted and categorised into low, medium or high empowerment. Given the specific marital context of certain items, SWPER included partnered women only for accuracy.

Statistical analysis

We conducted all the statistical analysis using R V.4.4.1. The help-seeking behaviour for symptoms of anxiety and depression of the respondents was summarised using descriptive statistics, such as proportions for categorical variables and means with SDs for continuous variables.

To explore the association between help-seeking behaviour and independent variables, we employed logistic regression models. In model 1, we applied bivariable logistic regression analyses to examine each independent variable individually to identify preliminary associations. Model 2 incorporated adjustments for sociodemographic variables, including area of residence (urban or rural), wealth index (poor, middle or rich), educational attainment (no formal education, primary, secondary or higher secondary/above), age (<30 years or >30 years) and employment status (employed/not employed). In our final model (model 3), we integrated the domains of the SWPER index alongside the sociodemographic variables. However, we excluded educational attainment and employment status from this model since these variables were already represented within the SWPER index, thus preventing multicollinearity. From the logistic regression models, we estimated adjusted ORs (aORs) with corresponding 95% CIs. We conducted model comparison using the Akaike information criterion (AIC). Both models demonstrated similar AIC values (approximately 2741), indicating comparable overall fit when accounting for model complexity. However, Nagelkerke’s R² was higher for model 3 (0.034) compared with model 2 (0.017), suggesting that model 3 explained a greater proportion of the variance in help-seeking behaviour. We considered a two-sided p<0.05 as significance threshold for statistical inference.

Findings

Sociodemographic characteristics

The mean age of the women was 33.3 years (SD 8.7 years). More than two-thirds of the women resided in rural areas. Around 40% of them were in the poorest or poor, 20% in the middle and 35% in the rich or richest quintile. Almost half of the women reached secondary or higher levels of education, while 30% were employed.

In the SWPER index, 83% of women were categorised as having high empowerment in the ‘attitude to violence’ domain, while the proportion of high empowerment was 16% in the ‘social independence’ domain. In the ‘decision-making’ domain, 60% of women demonstrated high empowerment (table 1).

Table 1. Sociodemographic characteristics of the women (n=2881).

Characteristics	n (%)
Residence
Urban	926 (32.14)
Rural	1955 (67.86)
Wealth quintile
Poorest	644 (22.35)
Poorer	585 (20.31)
Middle	617 (21.42)
Richer	594 (20.62)
Richest	441 (15.31)
Educational attainment
No formal education	506 (17.56)
Primary	835 (28.98)
Secondary	1225 (42.52)
Higher secondary/above	315 (10.93)
Age (years), mean (SD)	33.26 (8.69)
Age category (years)
<30	1142 (39.64)
≥30	1739 (60.36)
Current employment status
Unemployed	2007 (69.66)
Employed	874 (30.34)
Domains of women empowerment (SWPER index)
Attitude to violence
Low	134 (4.65)
Medium	347 (12.04)
High	2400 (83.30)
Social independence
Low	1258 (43.67)
Medium	1159 (40.23)
High	464 (16.11)
Decision-making
Low	469 (16.28)
Medium	666 (23.12)
High	1746 (60.60)

SWPER, Survey-based Women’s Empowerment.

Help-seeking behaviour

Among the women reporting symptoms of anxiety or depression, 18.7% (95% CI 17.3% to 20.2%) sought help for these symptoms, while approximately 5.2% (95% CI 4.5% to 6.1%) were taking medication. Of those who were taking medication, 2.7% (95% CI 1.9% to 3.5%) did so after seeking care from qualified healthcare providers (figure 1).

Figure 1. Help-seeking behaviour among women with symptoms of anxiety or depression, including those who sought help and those who did not (n=2881).

The majority of women who sought help for symptoms of anxiety or depression turned to their spouses or partners (43%, 95% CI 38.6% to 47.2%) or other family members (52%, 95% CI 47.8% to 56.4%) for support. Additionally, 23% (95% CI 19.6% to 26.8%) sought help from neighbours, and 8% (95% CI 5.7% to 10.5%) from friends. A smaller segment, around 15% (95% CI 11.8% to 18.0%), sought help from doctors or medical personnel (figure 2).

Figure 2. Sources of help-seeking among women with symptoms of anxiety or depression who sought help, categorised into formal (eg, doctor/medical personnel) and informal (eg, spouse, family, friends etc) sources (n=539).

Factors associated with help-seeking behaviour

In the multivariable logistic regression model including the sociodemographic variables (model 2), women aged 30 years or older had higher odds of seeking help for symptoms of anxiety and depression (aOR 1.25, 95% CI 1.02 to 1.53). Besides, higher odds of help-seeking were also observed among women who were in the rich wealth index group (aOR 1.49, 95% CI 1.09 to 1.89), those with secondary education (aOR 1.47, 95% CI 1.08 to 2.01) or higher (aOR 1.56, 95% CI 1.03 to 2.36), and those who were employed (aOR 1.46, 95% CI 1.19 to 1.78).

However, in the logistic regression model including SWPER index of women empowerment (model 3), high empowerment in the ‘attitude to violence’ domain was inversely associated with help-seeking behaviour (aOR 0.54, 95% CI 0.37 to 0.81), as was high empowerment in the ‘decision-making’ domain (aOR 0.53, 95% CI 0.41 to 0.68). No significant association was found between help-seeking behaviour and ‘social independence’ domain (table 2).

Table 2. Factors associated with help-seeking behaviour symptoms of anxiety or depression.

Variables	Model 1*	Model 2†	Model 3‡
	cOR (95% CI)	aOR (95% CI)	aOR (95% CI)
Age (years)
<30	Ref.	Ref.	Ref.
≥30	1.25 (1.03 to 1.52)	1.25 (1.02 to 1.53)	1.27 (1.04 to 1.55)
Residence
Urban	Ref.	Ref.	Ref.
Rural	1.06 (0.86 to 1.30)	1.06 (0.86 to 1.31)	1.01 (0.82 to 1.24)
Wealth index
Poor	Ref.	Ref.	Ref.
Middle	1.16 (0.87 to 1.55)	1.10 (0.82 to 1.49)	1.19 (0.88 to 1.59)
Rich	1.42 (1.04 to 1.92)	1.49 (1.09 to 1.89)	1.51 (1.10 to 2.07)
Educational attainment
No formal education	Ref.	Ref.
Primary	1.14 (0.85 to 1.54)	1.29 (0.96 to 1.76)
Secondary	1.22 (0.93 to 1.62)	1.47 (1.08 to 2.01)
Higher secondary/above	1.35 (0.94 to 1.93)	1.56 (1.03 to 2.36)
Current employment status
Unemployed	Ref.	Ref.
Employed	1.45 (1.19 to 1.76)	1.46 (1.19 to 1.78)
Women empowerment (SWPER index)
Attitude to violence
Low	Ref.		Ref.
Medium	0.60 (0.39 to 0.96)		0.64 (0.37 to 1.01)
High	0.51 (0.35 to 0.76)		0.54 (0.37 to 0.81)
Social independence
Low	Ref.		Ref.
Medium	0.84 (0.68 to 1.03)		0.85 (0.69 to 1.05)
High	0.93 (0.70 to 1.21)		0.88 (0.66 to 1.17)
Decision-making
Low	Ref.		Ref.
Medium	0.90 (0.69 to 1.19)		0.88 (0.66 to 1.16)
High	0.55 (0.43 to 0.70)		0.53 (0.41 to 0.68)

^*Unadjusted model.

^†Adjusted for sociodemographic variables.

^‡Adjusted for domains of SWPER index.

aOR, adjusted OR; cOR, crude OR; SWPER, Survey-based Women’s Empowerment.

Discussion

This study examined help-seeking behaviours for anxiety and depression symptoms among women in Bangladesh, along with factors associated with these behaviours. Findings showed that only 19% of women experiencing these symptoms sought help, with most turning to informal sources like family members, friends or neighbours. A smaller percentage, approximately 15%, accessed care from medical professionals. The analysis revealed positive associations between factors such as age, household wealth, educational level and employment status and help-seeking for mental health symptoms. In contrast, two women empowerment dimensions, ‘attitude towards violence’ and ‘decision-making’, were negatively associated with help-seeking, while ‘social independence’ showed no significant association.

The low rate of help-seeking for mental health symptoms among women, we observed in this study, highlights a considerable gap in their access to mental healthcare services. This finding aligns with the National Mental Health Survey⁴, which reported that only about 10% of women with mental health symptoms sought help, predominantly relying on informal support systems.⁴ This limited utilisation of formal healthcare services pointed to the insufficiency of structured mental health services in Bangladesh, where a few healthcare workers (0.49%) are trained in mental health and only 1.6 psychiatrists are available per one million people.⁵ The reliance on informal sources of support, while culturally significant, is insufficient for addressing mental health symptoms that may require professional intervention. Moreover, we found that nearly half of the women who were taking medication for anxiety and depression did so without seeking care from qualified healthcare providers. This practice may promote irrational use of these drugs which can lead to adverse health consequences such as drug dependency, inappropriate dosages or adverse drug interactions. To address this gap in utilisation of formal mental health services, the National Mental Health Policy of Bangladesh (2019) and National Mental Health Strategic Plan (2020–2030) emphasised integrating mental health into primary healthcare, expanding community-based services, training non-specialist providers and improving access to psychotropic medications at lower-tier facilities.^{24 25} In parallel, WHO’s Mental Health Gap Action Programme advocates for task-shifting, early intervention and culturally adapted community outreach to strengthen mental health systems in resource-constrained settings like Bangladesh.²⁶ However, the success of these initiatives fundamentally depends on community-level awareness, perceived need and health-seeking behaviours for mental health symptoms to drive effective utilisation of services.

One of our primary objectives was to explore the association between women empowerment and their help-seeking behaviours for mental health symptoms. To assess the women empowerment status, we used the SWPER index, a multidimensional measure derived from 62 DHS, including three empowerment domains: ‘attitude to violence’, ‘decision-making’ and ‘social independence’. We found a negative association between high empowerment status in the ‘attitude to violence’ and ‘decision-making’ domains and help-seeking behaviour for mental health issues among our respondents. Specifically, women who scored higher in these empowerment domains were less likely to seek help for their symptoms of anxiety and depression. It is possible that women with greater decision-making power or less tolerance for violence develop stronger self-reliance or prefer self-management strategies over formal care, especially in settings where mental health systems are weak. Additionally, distrust in the quality or confidentiality of formal care may dissuade these women from seeking professional help.²⁷

However, two core components of the SWPER index, such as educational attainment and employment status, were independently associated with increased odds of seeking help for mental health issues in our study. This seemingly contradictory finding may reflect limitations in the conceptualisation and scope of the SWPER index itself. While the index includes important domains of empowerment related to attitudes toward violence, decision-making and social independence, it does not fully accommodate other dimensions such as economic empowerment, asset ownership or contextual cultural factors that might influence mental health behaviours. For instance, higher scores in ‘attitude to violence’ and ‘decision-making’ may indicate assertiveness or resistance to traditional gender roles, but these qualities do not necessarily translate into increased utilisation of formal mental health services, especially in a setting where stigma, limited service availability and sociocultural norms strongly influence help-seeking.^{9 10} These limitations suggest that composite empowerment measures like SWPER may mask nuanced relationships between specific empowerment facets and mental health behaviours, indicating the need for more comprehensive and context-specific tools to assess the association between women empowerment status and mental health help-seeking in Bangladesh. However, this association may vary in different cultural and societal norms. For instance, a study in Nepal found that higher empowerment in the SWPER ‘decision-making’ domain was positively associated with mental healthcare-seeking.¹⁶

Besides, we acknowledge that the limitations of the SWPER index could be overcome by using other multidimensional indices of empowerment, such as the Women’s Empowerment in Agriculture Index²⁸ or other country-specific indices could incorporate broader aspects of empowerment. However, these are not available or compatible with DHS data structures. SWPER was selected because it is one of the most commonly used validated measures directly derived from DHS variables, allowing for population-level inference. This index has demonstrated utility in predicting healthcare utilisation behaviours, particularly in the domains of maternal and reproductive health. Previous studies have shown that higher empowerment scores, especially in the decision-making and social independence domains, are associated with increased use of antenatal care, institutional delivery and modern contraceptive methods in Bangladesh.^{17 29} These findings warrant further research, particularly qualitative or mixed-methods studies, to explore how empowered women perceive and respond to mental health needs. Existing literature offers limited qualitative evidence in this context. However, a social media content analysis study indicated that many women did not access mental health services despite exhibiting visible signs of psychological distress. Major barriers identified included low mental health literacy, stigma associated with help-seeking and limited availability of mental health services.²⁷

We found that educational attainment, a key indicator of empowerment, was independently associated with a higher likelihood of seeking mental healthcare. This finding aligns with previous studies indicating that higher education levels are associated with help-seeking for mental health issues.^{4 30} Higher educational attainment might increase awareness and reduce stigma about mental health issues.⁹ Besides, women with greater educational attainment might be more informed about the importance of mental health and be better equipped to navigate mental healthcare options.

Employment status, another dimension of empowerment, was also significantly associated with increased help-seeking for mental health symptoms among women. Employment may be associated with financial independence and personal agency, encouraging women to take proactive measures regarding their health and enabling them to make healthcare decisions without financial constraints or dependence on others. Although the ‘social independence’ domain of the SWPER index was not significantly associated with mental healthcare-seeking behaviour in our study, employment, as a distinct factor, likely facilitates increased exposure to mental health information and reduces the stigma associated with seeking care.³¹

Other significant associated factors with help-seeking behaviour among women for mental health symptoms identified in our study included age and wealth quintile. Older women demonstrated a higher likelihood of seeking support, possibly due to increased autonomy in healthcare decisions. Additionally, we found that women from wealthier households were more likely to seek help for their mental health symptoms. The higher wealth quintile is universally associated with better healthcare access, and mental healthcare service would not be an exception.³²

One of the major strengths of our study is we used a nationally representative dataset (BDHS 2022) to explore the relationship between multidimensional women’s empowerment and help-seeking behaviour for depression and anxiety symptoms, an area with limited empirical investigation, particularly in the context of Bangladesh. To our knowledge, this is among the first studies in Bangladesh to use the globally validated SWPER Global Index to assess women’s empowerment across multiple domains and examine its association with mental health service utilisation. The large, population-based sample enhances the generalisability of our findings to reproductive-aged women in Bangladesh, while the use of validated screening tools (PHQ-9 and GAD-7) ensures methodological rigour in identifying individuals with mental health symptoms. Furthermore, by integrating empowerment metrics with mental health and care-seeking variables, our study provides a unique and policy-relevant perspective on structural and social determinants influencing access to mental healthcare among women in Bangladesh.

However, our study had several limitations. First, we used the SWPER index to measure women empowerment status. Despite being a globally validated measure,²³ this index cannot measure empowerment as a multifaceted process involving economic, sociocultural, familial, interpersonal, legal, political, societal and psychosocial aspects.³³ Besides, the index lacks coverage of economic participation, asset ownership and governance involvement of women, limiting its effectiveness in representing empowerment comprehensively.²³ Additionally, the SWPER index is limited for married or cohabiting women; hence, we had to exclude unmarried or non-cohabiting women, who might be more vulnerable to mental health issues. This exclusion, particularly of unmarried women aged 15–19, may under-represent an important subgroup with distinct mental health needs and should be considered a key limitation of the study’s generalisability. We also excluded individuals with missing data in the mental health domain, which could have introduced selection bias if the missingness was non-random. Moreover, the time frame for assessing help-seeking behaviour was not explicitly defined in the BDHS questionnaire, which may have led to inconsistencies between the symptom reference period (past 2 weeks) and reported help-seeking, potentially affecting the accuracy of the estimated treatment gap. Additionally, we acknowledge the potential for non-response or social desirability bias in self-reporting of mental health symptoms and help-seeking behaviour. Furthermore, the relatively low Nagelkerke’s R² values observed in our models indicate that a substantial proportion of the variability in help-seeking behaviour remains unexplained, highlighting the influence of unmeasured factors not included in our analysis. One important factor we were unable to measure, but which is strongly supported in existing literature, is the perceived need for mental health treatment. Epidemiological studies consistently show that one of the most important reasons individuals do not seek mental healthcare is that they do not perceive a need for it, even when symptomatic.³⁴ The absence of data on perceived need in the BDHS survey limited our ability to distinguish between women who did not seek care due to lack of perceived need vs those who faced structural or social barriers despite recognising the need. This distinction is essential for interpreting treatment gaps and designing appropriate interventions, and our inability to account for this limits the depth of our conclusions. Finally, in a developing country like Bangladesh, people encounter several structural and social barriers to accessing and using mental health services including limited resources. Although our study indicated that women empowerment may not have a direct impact on mental health care-seeking behaviour, other elements, such as health literacy, sociocultural support and available healthcare resources, also shape mental health access. However, we were unable to explore these dimensions within this study.

Conclusions

In Bangladesh, only a small proportion of women with anxiety and depression symptoms actively seek mental health support, primarily from informal sources. Significant associated factors of help-seeking behaviour include age, education level, household wealth and employment status. Interestingly, while high empowerment level in attitude to violence and decision-making domain of SWPER index was negatively associated with help-seeking for mental health symptoms, specific factors like education and employment were positively associated with seeking care.

Data availability statement

Data are available in a public, open access repository.