STRENGTHS AND LIMITATIONS OF THIS STUDY

The strengths of this research include valuable insights into food security and mental health among under-represented urban slum dwellers in Bangladesh measured using validated tools (Household Food Insecurity Assessment Scale and Depression, Anxiety and Stress Scale-21) to ensure reliability.

Multivariable logistic regression model controls confounders, strengthening the findings on the link between food insecurity and mental health.

The use of a relatively large sample from a vulnerable group strengthens the relevance of the findings for policymakers targeting similar populations.

Cross-sectional design limits the ability to draw causal inferences between food insecurity and mental health outcomes.

The use of convenience sampling, limited resources and time constraints prevented random sampling, which may affect the generalisability of the findings.

Introduction

Bangladesh, known for its very high population density, is experiencing rapid urban influx. According to a recent statistic, around one-third of the population lives in the country’s urban areas.1 In urban city centres, the demand for housing often outstrips the available resources and infrastructure, leading to the emergence of slum settlements.2 In the context of urban settings, slum dwellers, with their marginalised status and substandard living conditions, epitomise a reflection of broader socioeconomic challenges.3 With limited access to essential services such as clean water, appropriate sanitation and healthcare, the high population density in these regions exacerbates existing socioeconomic disparities.3 4

The global economy has recently faced significant challenges, including economic instability and volatility.5 Economic hardship has had far-reaching effects, particularly on vulnerable populations in developing countries like Bangladesh. In addition to global economic uncertainty, urban slum communities in Bangladesh confront compounded difficulties due to soaring inflation and price hikes.6 Fluctuations in commodity prices, disruptions in supply chains, and shifts in consumer demand patterns have contributed to inflationary pressures, driving up the cost of living worldwide.7

In Bangladesh, where a substantial portion of the population lives below the poverty line, the impact of these economic dynamics is keenly felt. Rising inflation has eroded the purchasing power of slum residents, rendering essential goods and services increasingly unaffordable.8 This price escalation has exacerbated socioeconomic inequalities, pushing vulnerable communities further into hardship.8 Bare necessities such as food, shelter and healthcare have become out of reach for many, deepening poverty and deprivation within urban slum settlements.9 The ongoing inflationary pressures have significantly disrupted the economic recovery of low-income households in Bangladesh from the COVID-19 shock.9 As a result, many households have drastically reduced or stopped consuming major food items such as fish, meat, milk and fruit.10 This reduction in food consumption quantity and quality has been more pronounced in urban slums than in rural areas.11

Mental health is an essential dimension of overall well-being, yet it is frequently overlooked in the context of slum populations.12 Existing literature suggests that inflation-related hardships may exacerbate symptoms of depression and anxiety, particularly within vulnerable populations such as those residing in urban slums.13 Additionally, food insecurity has been linked to poor mental health, with research indicating a relationship between the intensity of household food insecurity and poor mental health outcomes.14 In fact, there is evidence of a dose–response relationship between the intensity of household food insecurity and poor mental health status, with episodes of severe food insecurity being characterised as a very stressful chronic condition.15 Overcrowding, inadequate sanitation and exposure to violence significantly contribute to the burden of mental health issues.16 According to Fang et al’s study, food insecurity is connected with a 257% greater risk of anxiety and a 253% higher risk of depression.17 Several studies have also examined the mental health impact of the Great Recession4 18 examining the repercussions of recession-induced adversities, such as reduced spending, heightened credit debt, missed rental payments, relocation to family homes and job loss, reveals a correlation between increased exposure to these hardships and elevated levels of psychological distress.19

The motivation for this research stems from the need to address the challenges slum populations face during recent economic stressors, which is intended to provide valuable insights. This study aims to explore the challenges faced by urban slum dwellers in Bangladesh during the current inflationary period, as well as explore the interrelationships between sociodemographic factors, food security and mental health among these vulnerable adults in the slum.

Methods and materialsStudy design and setting

A cross-sectional study was conducted on 300 slum dwellers aged 18–60 years between October 2023 and January 2024 in ‘Korail Basti’, an urban slum in Dhaka North City Corporation, Bangladesh. Korail slum spans certain Banani and Gulshan regions, covering the 19th and 20th wards of Dhaka City Corporation, encompassing approximately 99 acres of land, residing an estimated 100 000 people.20

Sample size and sample selection procedure

The study involved 300 residents from the ‘Korail Basti’ slum area, selected through a convenient sampling technique applied to slum households. Because the slum environment is so different from other places, convenient sampling was chosen for this study. Random selection was not achievable since many household members had moved frequently, and many more were frequently absent during the day for work. The fact that there were a lot of families living in one house made random selection even harder. So, convenient sampling ensured that the needed sample size could be reached in a practical and effective way. Although convenience sampling was employed due to the practical challenges of random selection, efforts were made to enhance the degree of representativeness by including participants from different households across different sections of the slum, ensuring diverse sociodemographic characteristics within the sample. Following the inclusion criteria, the information was collected from the head of the household. When the household head was unavailable, the spouse of the household head was interviewed. Participants aged 18–60 years who had been residing in the Korail Basti slum area for at least 6 months and were the head of the household or, in their absence, the spouse of the household head and were willing to provide informed consent and willing to provide informed consent were included. Individuals who had been residing in the slum area for less than 6 months had severe cognitive or communication impairments, were not residents of the Korail Basti slum area, such as visitors or temporary workers or declined to provide informed consent or chose to withdraw from the study at any point were excluded.

The sample size was calculated using the formula of Cochran’s,

With a 5% margin of error (e), considering the prevalence of mental disorder among slum people (p=23.00%) as reported in a similar study in India,21 the standard normal variate of 1.96 (z), the required sample size was 272. However, we reached 50 more samples (322 samples) than the sample size required to account for non-responses and incomplete interviews, ensuring the final sample would not be less than the calculated one. After carefully reviewing the data, 22 cases were excluded from the analysis due to various reasons, including non-response and non-participation, incomplete data, missing data and the presence of statistical outliers. Finally, information from the 300 respondents was gathered and included in the final analysis. A study flow diagram (figure 1) has been included to illustrate the participant recruitment process, eligibility screening and final inclusion in the analysis.

Figure 1

Flow chart of the participation selection process.

Data collection procedure

A semistructured questionnaire was developed using the Kobo Toolbox, and the data were collected using a face-to-face interview. The questionnaire was developed based on previously published research, with slight adjustments for the study cohort.22 23 Five trained data collectors were appointed for the data collection purposes. The interview process involved administering the Household Food Insecurity Scale and the Depression, Anxiety and Stress Scale-21 (DASS-21) questionnaire to assess food security and mental health status. Special attention was given to ensuring the privacy and comfort of participants during the data collection process, and steps were taken to minimise social desirability bias.

Measures

Participants completed a questionnaire comprising sociodemographic information, validated versions of the Household Food Insecurity Assessment Scale-HFIAS, and the 21-item DASS-21 scale.

Sociodemographic characteristics

Sociodemographic data were collected, including age, gender, marital status, duration of stay in the slum, number of children, number of sources of income, education level, occupation and working hours. We also included their income categories and asked whether they received subsidies from the government or private/NGO (Non-Governmental Organization) authorities.

Household Food Insecurity Scale

The HFIAS developed by FAO-FANTA was used to measure access to household food insecurity. HFIAS is a modified version of the method used to determine the prevalence of household food insecurity and identify shifts in the population’s access to food situation.24 Two types of questions are included in the HFIAS. The first form of question is known as an occurrence question. Nine questions about occurrences ask if a situation linked to food insecurity has ever happened in the last 4 weeks (30 days). After each question about intensity, there is a question about how often a condition was reported to have happened in the last 4 weeks.25 First, we divided households into four categories based on their food availability: secure, mild, moderate and severely insecure. Further, we combined ‘food-secured’ and ‘mild food insecurity’ into a single group (food-secured to mild food insecurity) because tiny factions were seen to be ‘food-secured’.

Depression, Anxiety and Stress Scale-21

The DASS-21 is a valid and reliable measure for assessing psychological health.26 DASS-21 is a condensed version of the 42-item DASS scale, which measures depression, anxiety and stress.27 There are seven items in each of the three subscales of the DASS-21 scale. This famous and extensively used DASS-21 scale has been translated, validated and used in the Bengali language previously.18 28 Respondents were asked to rate their mental distress over the previous 4 weeks on a 4-point Likert scale ranging from 0 (never) to 3 (almost always). The individual scores for depression, anxiety and stress were calculated by adding the scores for each of their seven items. The final score for each of the three dimensions was then multiplied by two to produce a score ranging from 0 to 42.29 Individual scores for these three subscales were then classified into five severity categories: normal, mild, moderate, severe and extremely severe.30

Patient and public involvement

We did not involve patients directly in the research process; however, we engaged the public at the initial stages to help shape the research questions and outcome measures. Their priorities, experiences and preferences influenced the focus of the study. The public was also consulted during the study design to ensure the methods aligned with community needs and concerns. To disseminate the study results, we plan to share the findings with the public and relevant communities, considering their preferences for the format and timing of the information to be shared. We will involve them in decisions regarding how to communicate the results in a way that is accessible and meaningful.

Statistical analysis

STATA V.16 was used for data analysis in this study. Outliers, identified as extreme values and missing data were removed to ensure the accuracy of our analysis. Outliers were identified using the IQR method, where values beyond 1.5 times the IQR from the first or third quartile were considered extreme and removed to minimise skewness in the dataset. We used list-wise deletion for missing data, removing cases with incomplete responses to maintain data integrity and ensure consistent sample size across analyses. In our analysis, list-wise deletion was used because the amount of missing data was minimal (less than 5%), and we assumed the data were missing completely at random. We conducted Little’s χ2 test for MCAR, which provided a p value of 0.179, further confirming our assumption. Therefore, we restrict ourselves from exploring alternatives, such as multiple imputations. Frequencies and percentages for each categorical variable are displayed. To evaluate the association between mental health status (depression, anxiety and stress) and other covariates, we tested three separate bivariate analyses (χ2 test). For the multivariable analysis, we used a binary logistic regression model to identify the factors associated with mental health status (stress, anxiety and depression) among adults in slums. We included all explanatory variables in the multivariable model based on theoretical relevance and to control for potential confounders.31

Regardless of their significance in the bivariate analysis, we included all variables in the model to minimise omitted variable bias. The covariates included gender, age, duration of staying in the slum, number of children, number of sources of income, education, occupation, income categories, subsidies provided by the Government/NGO and household food security status. ORs and 95% CIs were derived from the regression output. All statistical tests were two tailed, and a p<0.05 was considered statistically significant.

Results

Table 1 presents the sociodemographic characteristics of individuals residing in urban slums in Dhaka city. Of the participants, 38% were male and 62% were female. The age group with the highest representation (33.7%) was aged 26–35. The majority of participants were married (83.3%) and had been living in the slum for more than 5 years (80%). Most households (72.3%) had one to three children. Most participants derived income from a single source (72.3%), and only 38.2% attended school for 6 years or more. Many participants worked in the transportation and garment industry, comprising 21% of the total. Participants earning below 10 000 Bangladeshi Taka monthly constituted 62.3% of the sample. Only 23% reported receiving subsidies from the government or NGOs. The food security status indicated that 39.3% experienced mild food insecurity, 22.6% moderate food insecurity and 38% severe food insecurity, highlighting significant concerns within the surveyed population.

Table 1

Sociodemographic characteristics of the respondents

Figure 2 illustrates the mental health status of the 300 participants assessed using DASS-21. The mean scores±SD were (15.7±9.4) for depression, (11.6±8.8) for anxiety and (15.8±8.0) for stress. The results reveal that a significant number of participants experienced mild to extremely severe levels of depression, anxiety and stress. Specifically, 73.3% of participants reported experiencing mild to extremely severe depression, 63.3% reported mild to severe anxiety levels and 54.7% reported mild to severe stress levels.

Figure 2

The distribution of mental health status among slum adult participants.

The bivariate association analysis revealed significant relationships between various background characteristics and mental health dimensions among the studied population, as displayed in online supplemental table 1. Females exhibit significantly higher rates of depression (p<0.001), anxiety (p<0.001) and stress (p=0.01). Age also played a role, particularly among younger adults aged 18–25 years, who reported elevated anxiety levels compared with older age groups (p=0.003). Marital status highlighted that being single correlated with increased depression (p=0.005) and anxiety (p=0.004). There are also noteworthy associations between education level and anxiety (p<0.001) and depression (p=0.001), with higher levels of education associated with reduced levels of both conditions. In addition, servants and housemaids demonstrated significantly higher depression (p<0.001) and anxiety (p<0.001) compared with other professions. Furthermore, receiving subsidies from the government or NGOs was associated with lower depression rates (p=0.003). Household food insecurity, ranging from moderate to severe, was consistently linked with elevated depression (p<0.001), anxiety (p<0.001) and stress (p=0.001) levels.

Online supplemental table 2 presents results from three binary logistic regression analyses focusing on the influence of food insecurity and various demographic and socioeconomic factors on depression, anxiety and stress outcomes. In contrast, online supplemental figure 1 illustrates these findings graphically through forest plots displaying the ORs. Key findings include significantly higher odds of anxiety among females (aOR 4.1, 95% CI 2.0 to 8.4) compared with males. Among individuals aged 46–60 years, there were significantly increased odds of anxiety (aOR (adjusted Odds Ratio) 4.5, 95% CI 1.3 to 15.5) compared with those aged 18–25 years. Single individuals had substantially higher odds of depression (aOR 8.7, 95% CI 1.5 to 50.0) compared with married individuals. Part-time workers had 97% lower odds of depression (95% CI 0.003 to 0.4) compared with housemaids, while transport drivers had 80% lower odds of anxiety (aOR 0.2, 95% CI 0.03 to 0.9) compared with housemaids. Individuals earning less than 10 000 takas monthly had approximately 14% higher odds of anxiety compared with those earning over 20 000 takas monthly. The study also highlights the influence of food security on mental health outcomes, showing that severe food insecurity was associated with higher odds of depression (aOR 2.8, 95% CI 1.2 to 6.7), anxiety (aOR 8.2, 95% CI 3.3 to 20.3) and stress (aOR 4.9, 95% CI 2.3 to 10.5) compared with mild food insecurity or food security.

We conducted diagnostic checks to meet our binary logistic regression assumptions. Multicollinearity was assessed using the variance inflation factor, with all values below 5 indicating no severe collinearity. The Box-Tidwell test confirmed the linearity of continuous predictors with the log odds of the outcome. Influential observations were examined using Cook’s Distance, with no cases exceeding the threshold of 1. Model fit was evaluated using the Hosmer-Lemeshow test (p>0.05), suggesting the model fits our data well. These results indicate that the logistic regression model assumptions were satisfied, ensuring the robustness of our findings.

Discussion

This community-based cross-sectional study explored the sociodemographic characteristics, food security status and mental health of a vulnerable population living in a slum during the recent economic instability. We discovered that the majority of the households experienced food insecurity. A study conducted in Bangladesh during the early stages of the COVID-19 pandemic reported a significant rise in food insecurity than our reported results, emphasising its severity as a major public health issue.32 Disruptions in food supply chains, income loss and mobility restrictions further worsened the situation, particularly for low-income households.32 33 The difference in findings can be attributed to the timing of our research during the current economic recession in Bangladesh and the COVID-19 pandemic, which created a scar on the vulnerable slum population. This finding also highlights a striking prevalence of mental distress among slum residents, aligning with the global trend of increased mental health challenges during recent years.34–37 Living conditions and social surroundings in urban slums, as studied in Varanasi, India, showed a significant impact on an individual’s mental health.38 Recent evidence also suggests individuals with disabilities, lower socioeconomic status and those in deprived areas are at higher risk of poor mental health, as highlighted by Hasan et al in Bangladeshi adults and Kirkbride et al on social determinants of mental health in global north countries.18 39

This research highlights a crucial relationship between sociodemographic factors and mental health outcomes, particularly emphasising the impact of economic instability on urban slum dwellers in Bangladesh. Gender was identified as a key determinant of anxiety levels, with females reporting significantly higher rates of anxiety. This finding aligns with prior research conducted in Nepal and India demonstrating gender-based disparities in mental health, which may be influenced by sociocultural roles, economic dependence and restricted access to support systems.38 40 This discrepancy can be attributed to a combination of factors. Traditional gender roles prevalent in these societies often place women in caregiving roles, increasing their anxiety levels.41 Economic disparities and societal expectations may compel women to bear the dual burden of domestic responsibilities and financial struggles, which can exacerbate their mental health vulnerability.41 In Western and industrialised countries, firmly established societal norms surrounding appearance, thinness and caregiving contribute to higher rates of anxiety among women.42 Similarly, in Mexico, traditional gender norms place significant pressure on women to prioritise the needs of others over their own, leading to elevated levels of stress and anxiety.43

Age emerged as another significant determinant of mental health outcomes in this study, particularly with regard to anxiety and stress levels. Older participants exhibited a higher propensity for experiencing anxiety compared with their younger counterparts. This observation aligns with previous research indicating that older individuals often face heightened psychological distress due to health concerns, familial responsibilities and financial insecurities.44 A study conducted across 33 countries, including the USA, Canada, the UK and several European and Asian nations, revealed that older adults experienced significant deleterious effects on their mental health.45 This implies a general pattern of increased stress susceptibility to ageing, likely due to growing responsibilities and life changes as people move from their early 20s to their 40s. However, different age groups face distinct stressors while younger adults often struggle with financial concerns, career pressures and social media-induced anxiety, older adults may indeed experience health-related worries and concerns about family and financial security.46

The study identified a significant association between marital status and depression, highlighting the heightened psychological burden experienced by unmarried individuals. This correlation may stem from increased feelings of social isolation, limited emotional support and financial insecurities commonly faced by those who are single. Research revealed unmarried individuals may also lack the protective effects of spousal support, which has been shown to buffer against stress and enhance emotional well-being.47 Additionally, the absence of a stable partnership may exacerbate feelings of loneliness and uncertainty, further contributing to depressive symptoms.46 These findings are consistent with prior research demonstrating that married individuals often report better mental health outcomes due to shared responsibilities, companionship and emotional stability.48 49

Occupational diversity significantly influences participants’ mental health conditions, with transport drivers and part-time casual workers showing lower levels of depression and anxiety. This may be attributed to the stabilising effects of steady employment on both financial and emotional well-being, crucial for mitigating mental health issues such as depression, anxiety and stress.50 Economic implications are likely to exacerbate existing disparities in mental health outcomes across income groups,51 highlighting a distinct correlation between income levels and anxiety. Lower-income individuals are particularly vulnerable to heightened anxiety compared with their higher-income counterparts,52 consistent with broader research linking socioeconomic disparities to mental health disparities. This observation aligns with existing literature suggesting that limited access to resources, chronic stress and barriers to healthcare and mental health services worsen psychological distress in lower-income populations.52

The findings of this research emphasise the substantial impact of food security on mental health in slum populations, consistent with previous research demonstrating adverse psychological effects of inadequate food access.32 53 Furthermore, global evidence also indicates a strong association between food insecurity and poorer mental health, with a dose–response relationship observed across 149 countries.54 Several factors contribute to the mechanism underlying this relationship. Anxiety and depression are two mental health conditions that can be made worse by food insecurity.54 55 Food insecurity is a common cause of chronic stress and anxiety because it is difficult to know how to meet basic nutritional needs.56 Evidence suggests malnutrition has an effect on cognitive processes and emotional regulation, which in turn worsens mental health outcomes.56 57 In addition, food insecurity is linked to socioeconomic vulnerabilities, such as poverty and unemployment, which leads to a volatile relationship in which psychological distress and financial instability reinforce one another.55 58 Additionally, the stigma and embarrassment associated with food insecurity can lead to social isolation, which is troubling to mental health because it limits access to emotional support and community resources.59 Contextual factors such as socioeconomic status and community support may mediate these relationships, with research indicating stronger associations between food insecurity and depression in low-income households.56 However, some studies have reported variations, such as no significant association between food insecurity and anxiety among older adults in slums, highlighting the need for further investigation into nuanced differences across diverse slum populations.39 60 These variations could be influenced by methodological differences in measuring food insecurity, mental health outcomes and demographic differences among study populations.61

Policy implication

This study underscores the need for targeted policies to address challenges slum populations face during inflationary pressures. Locally, interventions should include inflation-indexed food subsidies, economic empowerment programmes (eg, vocational training, fair wages) and improved access to mental health services, education and infrastructure like housing, water and sanitation. Policies must address age-specific vulnerabilities, support vulnerable occupational groups and promote work–life balance through affordable childcare and fair labour practices. Additionally, community-based mental health programmes, such as counselling services and peer support groups, should be integrated into existing food security initiatives to address the psychological impact of food insecurity. Globally, efforts should focus on stabilising food prices, increasing food security and mental health funding, integrating mental health into SDGs (Sustainable Development Goals), and promoting fair trade. International collaborations can facilitate knowledge sharing and resource allocation to support low-income countries in addressing these dual challenges. A coordinated local-global approach, and cross-sectoral partnerships involving governments, NGOs and private sectors, can build resilient, equitable communities for slum dwellers to thrive. Furthermore, policies should prioritise gender-sensitive interventions, as women are disproportionately affected by food insecurity and mental health challenges, and ensure that marginalised groups are included.

Strengths and limitations

The strengths of this research include valuable insights into food security and mental health among under-represented urban slum dwellers in Bangladesh, measured using validated tools (HFIAS and DASS-21) to ensure reliability. Multivariable logistic regression model controls confounders, strengthening the findings on the link between food insecurity and mental health. The study’s timeliness, addressing economic instability and COVID-19 impacts, makes these findings highly relevant for vulnerable populations in developing countries like Bangladesh. This study’s cross-sectional design limits causal inferences between food insecurity and mental health. Our study employed convenience sampling, which may introduce selection bias. We conducted sensitivity analyses to supplement the impact of potential biases due to the use of convenient sampling. We conducted gender-disaggregated logistic regression analyses to examine the association between food security and depression, adjusting for age, marital status, slum duration, number of children, income sources, education, occupation, income categories and income subsidies. Food moderate insecurity was significantly associated with higher odds of depression among both males (aOR 8.9, 95% CI 1.7 to 47.5) and females (aOR 3.1, 95% CI 1.7 to 11.2). On the other hand, in the overall model combining both genders, food insecurity also showed significantly higher odds of depression (aOR 4.0, 95% CI 1.1 to 16.2). These findings provide us with the same directional relationship between food insecurity and depression, although the strength of the association varies between males and females. We have obtained similar results for the other two dependent variables, anxiety and stress.

However, our sample’s sociodemographic characteristics align closely with those from studies using random or cluster sampling.12 62 63 For instance, the age distribution, marital status, gender distribution and education status in our study are comparable to those reported in prior studies of slum populations in Bangladesh and India. These similarities suggest that our sample, while not randomly selected, captures key population characteristics relevant to the study question. Furthermore, our findings on the association between food insecurity and mental health are consistent with studies using more rigorous sampling methodologies. The slightly higher prevalence of mental health issues in our study may be due to contextual factors such as post-COVID-19 economic instability and inflationary pressures. While these points strengthen the credibility of our findings, we acknowledge the limitations of convenience sampling and recommend that future studies use random sampling where feasible to enhance generalisability further.

Conclusions

In conclusion, this study emphasises the significant influence of sociodemographic factors, food security and occupational diversity on the mental well-being of slum residents during the recent economic downturn. Policy recommendations should prioritise improving economic opportunities, enhancing food access, promoting financial literacy and investing in educational initiatives to alleviate mental health burdens in vulnerable populations. Further research is needed to deepen our understanding of these relationships and inform more effective interventions.

Data availability statement

Data are available on reasonable request.

Ethics statementsPatient consent for publication

Consent obtained directly from patient(s).

Ethics approval

This study has received ethical approval from the Institutional Review Board of South Asia University (Reference number: MNFS31/EA23). Before participation, all individuals provided informed written consent, underscoring the voluntary nature of their involvement and the confidentiality of their data. Adhering to the ethical principles outlined in the Declaration of Helsinki, every aspect of data collection was conducted with utmost regard for the well-being and rights of the participants.

Acknowledgments

The authors thank Jannatul Fedoushi Mow and Md. Asaduzzaman for their assistance in data organisation.