Identifying Associated Factors of Common Mental Health Disorders in Riverbank Erosion Areas of Bangladesh Using Machine Learning Algorithms

Identifying Associated Factors of Common Mental Health Disorders in Riverbank Erosion Areas of Bangladesh Using Machine Learning Algorithms | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Identifying Associated Factors of Common Mental Health Disorders in Riverbank Erosion Areas of Bangladesh Using Machine Learning Algorithms Md. Tahidur Rahman, Md. Lotifur Rohman, Shyam Sundar Sarkar, Md. Fahim, and 2 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-9455372/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted 6 You are reading this latest preprint version Abstract Background Riverbank erosion is a recurrent environmental hazard in Bangladesh that leads to displacement, loss of livelihood, loss of land, and social instability. These disruptions place affected populations at heightened risk of common mental health disorders (CMHDs), including depression, anxiety, and stress. However, evidence on the relative importance of socioeconomic, environmental, and displacement-related determinants remains limited. Objective This study aimed to identify key determinants of common mental health disorders among individuals living in riverbank erosion areas of Bangladesh using machine learning algorithms. Methods We analyzed cross-sectional data collected from erosion-affected communities, where household heads were interviewed face-to-face between September 2021 and January 2022 to assess Common Mental Health Disorders (CMHDs) using the Depression, Anxiety, and Stress Scale (DASS-21) (Depression ≥ 10, Anxiety ≥ 8, or Stress ≥ 15). Descriptive statistics were conducted to determine the prevalence of CMHDs, while χ²-test and logistic regression (LR) analysis were applied to identify statistically significant risk factors associated with CMHDs. Moreover, three optimized machine learning (ML) algorithms- Random Forest (RF), XGBoost, and LR were implemented for predicting CMHDs. Results A total of 711 households were included, and randomly allocated to training (70%; n = 497) and testing (30%; n = 214) datasets. Among 497 respondents, 398 (80.08%) were exposed to riverbank erosion. The prevalence of common mental health disorders was significantly higher among the exposed group than the non-exposed group (exposed: 84.17% versus non- exposed: 15.83%, P < 0.001). The Random Forest classifier was found to be the best predictive model, with the accuracy (0.776) and precision (0.807), while loss of house, displacement, and loss of land were found to be the most influential predictors of CMHDs risk. Conclusion This study identified and predicted key risk factors for common mental health disorders among exposed individuals in riverbank erosion areas of Bangladesh using ML algorithms, which may assist policymakers in mitigating the burden of CMHDs, with particular attention to housing loss, displaced populations, and low-income individuals. Depression Anxiety Stress Riverbank erosion Loss of house Machine Learning Bangladesh Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 1. Introduction According to the World Health Organization (WHO), depression and anxiety are the two most Common Mental Health Disorders globally, affecting approximately 300 million people worldwide [ 1 ]. These disorders can impair daily functioning, lower work or academic performance, strain relationships, and, in severe cases, lead to suicide. At the societal level, they contribute to a significant economic burden [ 2 ]. The Depression Anxiety Stress Scales (DASS) were systematically developed based on the tripartite model of depression and anxiety to assess symptoms of Common Mental Health Disorders (CMHDs) [ 3 ]. Globally, it is estimated that about 280 million people suffer from depression, while around 272.5 million are affected by anxiety disorders [ 4 , 5 ]. Natural disasters affect millions of people each year, not only physically but also psychologically [ 6 ]. Among such disasters, riverbank erosion has been associated with increased rates of depression, anxiety, and stress in both developed and developing countries. Climate change has further intensified the frequency and severity of such extreme weather events, including riverbank erosion, which now occur beyond the range of normal natural variability, thereby disrupting social, economic, and ecological systems [ 7 , 8 ]. In South Asia, Bangladesh reports a higher prevalence of mental illness compared to neighbouring countries [ 9 ]. Among Bangladeshi adults, 16.8% of males and 17% of females experience mental health issues, yet 32.3% of them do not seek medical care [ 10 ]. Around 20 of the 64 districts in Bangladesh are highly vulnerable to riverbank erosion [ 11 ]. In 2019 alone, approximately 5,081 households were displaced, and nearly 2,270 hectares of land were lost due to erosion [ 12 , 13 ]. According to Khatun et al. [ 14 ], if natural disasters keep happening more often and become stronger, about 30,366,230 households in Bangladesh could be affected by 2030. These extreme weather events contribute not only to widespread displacement and forced migration but also to deteriorating conditions that influence both physical and mental well-being [ 15 ]. We conducted a comprehensive literature review and found no population-based study in Bangladesh that specifically investigates the prevalence and risk factors of CMHDs in river-prone areas. Globally, only a limited number of studies have explored the common mental health impacts in other populations, such as the prevalence and predictors of CMHDs among pregnant women in Malaysia and a rural community-based cohort of northern Vietnamese women [ 16 , 17 ] and the association between Common Mental Health Disorders and body weight in Eastern Saudi Arabia [ 18 ]. To the best of our knowledge, no prior research in Bangladesh or elsewhere has identified the key risk factors of Common Mental Health Disorders (CMHDs), based on Depression, Anxiety, and Stress (DAS) scores, among individuals affected by riverbank erosion. Therefore, this study aims to investigate how river erosion influences Common Mental Health Disorders in Bangladesh and to identify the risk factors associated with these conditions among affected populations. 2. Methods A. Study area and sampling procedure We purposefully chose three districts of Kushtia, Tangail, and Rajbari due to the high risk of river erosion. We initially selected three erosion-prone Upazila Kumarkhali in the Kushtia district, Bhuyapur Upazila in the Tangail district, and Rajbari Sadar Upazila in the Rajbari district. From each of these subdistricts, one union was selected: Shiladah Union from Kumarkhali, Arjuna Union from Bhuyapur, and Mijanpur Union from Rajbari Sadar. We then randomly selected seven villages (Ramkrisnopur, Charjoukuri, Ramcondropur, Chilimpur, Dhunchi, Sonakandor and Charnarayanpur) from Shiladah union, two villages (Kalua, Kumarkandi) from Arjuna union, and four villages (Arjuna, Tarai, Dhubliya, and Kuthiboira) from Mijanpur union. A total of 13 villages were chosen from the three districts. The detail description of the selected household sample is presented in Fig. 1 . We calculated the sample size using the formula for a two-stage cluster sampling design: Step 1: Calculate initial sample size ( \(\:{n}_{0}\) ) : $$\:{n}_{0}=\frac{{Z}^{2}P(1-P)}{{d}^{2}}$$ Where: \(\:Z=1.96\) (for 95% confidence), \(\:P=0.5\) (assumed prevalence), \(\:d=0.05\) (margin of error) [ 17 ] $$\:{n}_{0}=\frac{(1.96{)}^{2}\times\:0.5\times\:0.5}{(0.05{)}^{2}}=\frac{3.8416\times\:0.25}{0.0025}=\frac{0.9604}{0.0025}=384.16\approx\:384$$ So, the initial sample size \(\:{n}_{0}=384\) Step 2: Adjust for design effect ( \(\:Deff\) ) : $$\:Deff=1+(m-1)\rho\:$$ Where: \(\:m=50\) (average cluster size), \(\:\rho\:=0.01\) (intra-cluster correlation) $$\:Deff=1+(50-1)\times\:0.01=1+49\times\:0.01=1+0.49=1.49\approx\:1.50$$ Step 3: Apply design effect to get adjusted sample size ( \(\:{n}_{\text{deff}}\) ) $$\:{n}_{\text{deff}}={n}_{0}\times\:Deff=384\times\:1.50=576$$ So, the sample size after adjusting for the design effect is 576. Step 4: Adjust for non-response rate ( \(\:r\) ) $$\:{n}_{\text{final}}={n}_{\text{deff}}\times\:(1+r)=576\times\:(1+0.10)=576\times\:1.10=633.6\approx\:634$$ Therefore, the minimum required sample size was 634, and a total of 711 households were finally selected for the study to ensure greater precision, comprising 100 from Shiladha Union (Kushtia District), 201 from Arjuna Union (Tangail District), and 410 from Mijanpur Union (Rajbari district). B. Study variables Outcome variables : The outcome variable in this study is Common Mental Health Disorders (CMHD), defined based on three mental health indicators: depression, anxiety, and stress. These were assessed using the Depression Anxiety Stress Scales–21 items (DASS-21), which comprises three subscales—each with 7 items—corresponding to depression, anxiety, and stress. Participants responded to each item on a 4-point Likert scale ranging from 0 ("Did not apply to me at all") to 3 ("Applied to me very much or most of the time"). The total score for each subscale ranges from 0 to 21 and is multiplied by two (×2) to align with the DASS scoring guidelines. The following cut-off scores were used to determine the presence of CMHD symptoms: Depression: ≥10, Anxiety: ≥8, Stress: ≥15 [ 16 , 19 ]. We defined a binary variable, where "1" indicates the presence of Common Mental Health Disorders (CMHD), and "0" indicates their absence. C. Measures Depression, Anxiety and Stress Scales Depression Involves symptoms like sadness, loss of interest, sleep or appetite problems, guilt, trouble concentrating, and fatigue. It can be caused by genetic, biological, environmental, sociodemographic, and psychological factors [ 20 ]. Anxiety Anxiety is a reaction to perceived threats, marked by troubling thoughts, tension, fast heartbeat, sweating, dizziness, and chest pain. It may result from natural disasters, poverty, family issues, personality traits, or brain chemistry [ 21 ]. Stress arises when people feel overwhelmed by demands. Symptoms include mood changes, flashbacks, avoidance, and heightened alertness. Causes include trauma, loss, serious illness, or accidents [ 22 ]. Predictor variables : The predictor variables included socioeconomic, demographic, and riverbank erosion-related factors, as well as the household's exposure status. Riverbank erosion-related variables included the distance of the homestead from the river (in miles), time since last displacement, amount of agricultural land (yes/no), loss of agricultural land (yes/no), cattle loss (yes/no), death of a relative (yes/no), home loss (yes/no), social detachment (yes/no), substance misuse (yes/no), social encouragement for rehabilitation (yes/no), and hope for land recovery (yes/no). Demographic characteristics comprised sex (male/female), age group (in years), and number of children. Socioeconomic factors included the respondent's occupation, household income, and educational status. Displacement, a direct consequence of riverbank erosion, was categorized into three groups based on the time since the last displacement: non-displaced, displaced within the past three years, and displaced more than three years ago. Age was categorized into three groups: 22–37 years, 38–45 years, and 46–80 years. Similarly, the number of children was grouped into three categories: 1–2, 3–4, and more than 4. D. Feature Selection Feature selection was performed to identify the most relevant predictors for the machine learning models. In the first stage, a chi-square test was applied to identify variables significantly associated with the outcome variable, and only statistically significant variables held. In the second stage, logistic regression analysis was conducted on those chi-square significant variables to further assess their independent effect. A p-value < 0.05 was used to determine the statistical significance of these associations. These selected features were subsequently used for model development and evaluation on the testing dataset.. All statistical analyses were conducted using Stata version 15 (StataCorp, College Station, TX, USA) and Python (Jupyter Notebook; ipykernel 6.25.1). E. Machine Learning Framework and Algorithms The overview of the proposed machine-learning (ML) framework for predicting common mental health disorders (CMHD) is illustrated in Fig. 2 . First, the complete dataset (including participants with and without CMHD) is randomly divided into two subsets: 70% training data and 30% testing data. Using the training subset, a univariate feature screening step is performed in which each candidate predictor is evaluated against the CMHD outcome using the Chi-square test of independence, and p-values are computed for all features. Features meeting the predefined significance threshold (p < 0.05) are retained as the optimal feature set [ 23 ]. Next, the reduced training data (containing only the optimal features) is used to build an ML‑based prediction system by training three classifiers: Logistic Regression (LR), Random Forest (RF), and XGBoost Logistic Regression (LR): Utilized as a linear baseline to compare against ensemble methods [ 24 ]. Random Forest (RF): An ensemble bagging technique that builds several decision trees to manage feature interactions and increase classification accuracy [ 25 ]. XGBoost: A gradient boosting framework that builds sequential trees to minimize residual errors through an iterative optimization [ 26 ]. F. Hyperparameter Tuning and Selection Criteria Grid search [ 27 ] with fivefold cross-validation was used for systematic hyperparameter optimization to ensure the robustness and generalizability of the prediction models. By dividing the training data into five separate subsets and iteratively training the models on four folds while validating on the remaining one, this method made it possible to thoroughly evaluate different parameter combinations. This method guarantees that the chosen parameters function consistently across several data segments and reduces the possibility of overfitting [ 28 ]. The hyperparameters to be used in optimization are given in Table 1 . Table 1 Hyperparameter optimization of algorithms using the grid search method Algorithms Search range of each parameter LR 'C': [0.001, 0.01, 0.1, 1, 10, 100], 'penalty': ['10', '12','13', '14'], 'solver': ['lbfgs', 'liblinear'] RF 'n_estimators': [100, 200, 300, 400, 450, 500, 550, 600], 'max_depth': [None, 10, 20], 'min_samples_split': [ 2 , 5 , 8 , 10 ], 'max_features': ['sqrt', 'log2'] XGBoost 'n_estimators': [100, 200, 300, 400], 'max_depth': [ 3 , 5 , 7 ], 'learning_rate': [0.01, 0.1, 0.2], 'subsample': [0.8, 1.0], 'colsample_bytree': [0.8, 0.9, 1.0, 1.2] The model parameters were estimated through maximum likelihood estimation (MLE) using the training dataset. The fitted model was subsequently employed to predict class membership (CMHD vs. healthy controls) in the test dataset. Furthermore, class probabilities for the response variable were computed to quantify the likelihood of each predicted outcome. G. Model Evaluation and Feature Importance A wide range of evaluation metrics were used to evaluate the optimized models' performance: Accuracy: The model's overall accuracy [ 29 ]. The model's capacity to distinguish between CMHD and non-CMHD situations is measured by the Area under the Receiver Operating Characteristic Curve (AUC) [ 30 ]. F1-Score: The harmonic mean of recall and precision, guaranteeing a fair assessment for classification [ 31 ]. Recall: Particularly given top priority in order to identify as many susceptible people as possible [ 32 ]. 3. Results Data were collected from 711 households to assess common mental health conditions in river erosion–affected areas of Bangladesh. The dataset was randomly divided into a training set (70%; n = 497) and a testing set (30%; n = 214). All subsequent analyses were conducted using the training sample (n = 497). A. Bivariate analysis of CMHD and associated factors The largest proportion of participants was from the Rajbari district (n = 281; 56.54%), followed by Tangail (n = 141; 28.37%) and Kushtia (n = 75; 15.09%). Approximately 47.48% of respondents were uneducated, and 78.27% did not own any agricultural land. Of the total respondents, 397 (80.08%) were exposed to riverbank erosion, while 99 (19.92%) were not. Among the total participants, 6.64% reported the loss of cattle, 10.87% lost a relative, 44.27% lost their homestead, and 77.06% lost agricultural land due to erosion (Table 2 ). According to regional distribution, Rajbari had the highest percentage of respondents (56.54%), followed by Tangail (28.37%) and Kushtia (15.09%) (Table 2 ). Figure 3 illustrates that among the exposed respondents, the highest proportion (84.17%) exhibited abnormal mental health conditions, indicating a strong association between erosion exposure and mental disorders. The bivariate distribution of Common Mental Health Disorders (CMHD) across different predictors is presented in Table 2 . We observed that the proportion of CMHD was higher in Rajbari (80.43%) and Tangail (78.01%) than in Kushtia (57.33%), a difference that is statistically significant (p < 0.001). The prevalence of CMHD increased with age, peaking at 84.66% in the oldest age group (46–80 years; p = 0.007). Compared to their counterparts, respondents who were uneducated (80.83%), had more than four children (90.20%), did not own agricultural land (80.46%), and had lower monthly earnings ( > = 6000 $ taka; 82.93%) showed the highest rates of CMHD. These differences were statistically significant (p < 0.05).By profession, the highest rates of CMHD were found among day laborers (87.23%) and drivers (83.33%). Among respondents who were exposed and relocated within the last three years, the prevalence of CMHD was significantly higher (97.56%) compared to those who did not migrate (59.01%; p < 0.001). We found that the prevalence of CMHD was statistically higher among those who lost their cattle (93.94% vs. 75.00%, p = 0.023), houses (94.55% vs. 61.73%, p < 0.001), and agricultural land (84.86% vs. 47.37%, p < 0.001) than among those who did not experience such losses. Furthermore, respondents who required social encouragement (88.56% vs. 67.91%, p < 0.001), experienced social detachment (91.82% vs. 68.93%, p < 0.001), and had hope for land return (87.25% vs. 73.42%, p < 0.001) also exhibited significantly higher rates of CMHD. Table 2 Prevalence of common mental health disorders with different predictors. Characteristics Overall N = 497(100%) n (%) Common mental health disorders \(\:{\chi\:}^{2}\) -value p-value Normal N = 118 (23.74%) Abnormal N = 379 (76.26%) n (%) n (%) Socio-demographic and economic factors Exposed Yes 398 (80.08) 63 (15.83) 335 (84.17) 66.929 < 0.001 No 99 (19.92) 55 (55.56) 44 (44.44) Area Kushtia 75 (15.09) 32 (42.67) 43 (57.33) 17.773 \(\:<0.001\) Tangail 141 (28.37) 31 (21.99) 110 (78.01) Rajbari 281 (56.54) 55 (19.57) 226 (80.43) Age group(year) 22–37 170 (34.21) 45 (26.47) 125 (73.53) 9.825 0.007 38–45 164 (33.00) 48 (29.27) 116 (70.73) 46–80 163 (32.8) 25 (15.34) 138 (84.66) Educational status Uneducated 236 (47.48) 45 (19.07) 191 (80.83) 4.943 0.026 Educated 261 (52.52) 73 (27.97) 188 (72.03) Profession Housewife 230 (46.28) 59 (25.65) 171 (74.35) 9.709 0.137 Farmer 62 (12.47) 13 (20.97) 49 (79.03) Business 55 (11.07) 19 (34.55) 36 (65.45) Day labor 47 (9.46) 6 (12.77) 41 (87.23) Service 23 (4.63) 7 (30.43) 16 (69.57) Driver 30 (6.04) 5 (16.67) 25 (83.33) Others 50 (10.06) 9 (18.00) 41 (82.00) No of children 1–2 196 (39.44) 52 (26.53) 144 (73.47) 6.374 0.041 3–4 250 (50.30) 61 (24.40) 189 (75.60) > 4 51 (10.26) 5 (9.80) 46 (90.20) Monthly earnings (taka) ≤ 6000 41 (8.25) 7 (17.07) 34 (82.93) 13.752 0.003 6001–10000 157 (31.59) 26 (16.56) 131 (75.25) 10001–15000 202 (40.64) 50 (24.75) 152 (83.44) ≥ 15000 97 (19.52) 35 (36.08) 62 (63.92) Agricultural land No 389 (78.27) 76 (19.54) 313 (80.46) 16.432 < 0.001 Yes 108 (21.73) 42 (38.89) 66 (61.11) Amount of agricultural land (decimal) No agricultural land 389 (78.27) 76 (19.54) 313 (80.46) 18.43 < 0.001 1–66 77 (15.49) 28 (36.36) 49 (63.64) 67–99 11 (2.21) 5 (45.45) 6 (54.55) ≥ 100 20 (4.02) 9 (45.00) 11 (55.0) Factors contributing to riverbank erosion Exposure category Exposed 398 (80.08) 63 (15.83) 335 (84.17) 66.929 < 0.001 Non-exposed 99 (19.92) 55 (55.56) 44 (44.44) Duration of internal displacement (Years) Not migrate 222 (44.67) 91 (40.99) 131 (59.01) 78.148 3 111 (22.33) 23 (20.72) 88 (79.28) Distance of homestead from the river (in miles) ≤ 0.2 0.978 > 0.2 Loss of cattle No 464 (93.36) 116 (25.00) 384 (75) 5.103 0.023 Yes 33 (6.64) 2 (6.06) 31 (93.94) Death of relative No 443 (89.13) 333 (75.17) 110 (24.83) 0.143 Yes 54 (10.87) 46 (85.19) 8 (14.81) Loss of agricultural land No 144 (22.94) 60 (52.63) 54 (47.37) 66.136 < 0.001 Yes 383 (77.06) 58 (15.14) 325 (84.86) Substance misuse No 481 (96.78) 117 (24.32) 364 (75.68) 1.885 0.1698 Yes 16 (3.22) 1 (6.25) 15 (93.75) Social encouragement No 296 (59.56) 95 (32.09) 201 (67.91) 27.07 < 0.001 Yes 201 (40.44) 23 (11.44) 178 (88.56) Loss of home No 277 (55.73) 106 (38.27) 171 (61.73) 71.114 < 0.001 Yes 220 (44.27) 12 (5.45) 208 (94.55) Social detachment No 338 (68.01) 105 (31.07) 233 (68.93) 30.038 < 0.001 Yes 159 (31.99) 13 (8.18) 146 (91.82) Hope for land return No 395 (79.48) 105 (26.58) 290 (73.42) 7.826 < 0.001 Yes 102 (20.52) 13 (12.75) 89 (87.25) P-values for each variable were computed using the \(\:{\chi\:}^{2}\) test statistic. A Parcats Plot (Fig. 4 ) illustrates a comprehensive snapshot of the study and visualizes the interaction between the most significant determinants identified by the chi-square tests: Displacement, Exposure, Loss land, and Loss house. The concentration of CMHD was significantly higher among those who had frequent displacements (> 3 times), even if the percentage of people with lower displacement frequency ( < = 3 times) was higher. Participants who lost their homes showed the lowest levels of psychological resilience (green routes), while those who did not lose any assets showed the highest levels. Critical differences in stressors at the community, household, and environmental levels were further highlighted by the much greater trajectories toward CMHD of those who experienced the junction of high exposure and dual land-house loss compared to those who did not. B. Binary Logistic Regression Analysis of CMHD and Associated Factors The logistic regression analysis's findings are shown in Table 3 . Higher probabilities of aberrant outcomes were highly correlated with exposure (OR = 6.65, 95% CI: 4.12–10.73; p < 0.001). The probabilities of aberrant outcomes were considerably greater for participants who lost agricultural land (OR = 6.23, 95% CI: 3.92–9.88; p < 0.001), livestock (OR = 5.17, 95% CI: 1.22–21.92; p = 0.026), and home (OR = 10.74, 95% CI: 5.72–20.18; p < 0.001). Conversely, having formal education (OR = 0.61, 95% CI: 0.40–0.93; p = 0.020), having a larger monthly income (≥ 15000 taka: OR = 0.36, 95% CI: 0.15–0.91; p = 0.030), owning agricultural land (OR = 0.38, 95% CI: 0.24–0.61; p < 0.001). In addition to psychosocial characteristics including social encouragement (OR = 3.66, 95% CI: 2.22–6.02; p < 0.001) and social detachment (OR = 5.06, 95% CI: 2.74–9.34; p < 0.001), respondents with more than four children had greater odds (OR = 3.32, 95% CI: 1.25–8.82; p = 0.016). Geographically, respondents from Rajbari (OR = 1.90, 95% CI: 1.11–3.25; p = 0.019) and Tangail (OR = 2.48, 95% CI: 1.33–4.62; p = 0.004) had greater odds than those from Kushtia. Increased risks of aberrant outcomes were also substantially correlated with older age (46–80 years: OR = 1.99, 95% CI: 1.15–3.43; p = 0.014) and internal displacement (≤ 3 years: OR = 27.79, 95% CI: 9.94–77.64; p 3 years: OR = 2.66, 95% CI: 1.56–4.52; p < 0.001). Table 3 Risk Factors Associated with Common Mental Health Disorders: Binary Logistic Regression Analysis. Characteristic OR (95% CI) p-value Agricultural land No (Ref.) 1 Yes 0.38 (0.24–0.61) < 0.001 Exposed No (Ref.) 1 Yes 6.65 (4.12–10.73) < 0.001 Loss of agricultural land No (Ref.) 1 Yes 6.23 (3.92–9.88) < 0.001 Loss of cattle No (Ref.) 1 Yes 5.17 (1.22–21.92) 0.026 Monthly earnings (taka) ≤ 6000 1 6001–10000 1.04 (0.42–2.59) 0.937 10001–15000 0.63 (0.26–1.50) 0.293 ≥ 15000 0.36 (0.15–0.91) 0.03 Loss of home No (Ref.) 1 Yes 10.74 (5.72–20.18) 4 3.32 (1.25–8.82) 0.016 Social encouragement No (Ref.) 1 Yes 3.66 (2.22–6.02) < 0.001 Social detachment No (Ref.) 1 Yes 5.06 (2.74–9.34) < 0.001 Area Kushtia (Ref.) 1 Tangail 2.48 (1.33–4.62) 0.004 Rajbari 1.90 (1.11–3.25) 0.019 Amount of agricultural land (decimal) No agricultural land (Ref.) 1 1–66 0.42 (0.25–0.72) 0.001 67–99 0.29 (0.09–0.98) 0.046 > 100 0.30 (0.12–0.74) 0.009 Educational status Uneducated (Ref.) 1 Educated 0.61 (0.40–0.93) 0.02 Age group (year) 22–37 (Ref.) 1 38–45 0.87 (0.54–1.40) 0.569 46–80 1.99 (1.15–3.43) 0.014 Duration of internal displacement (Years) Not migrate 1 ≤ 3 27.79 (9.94–77.64) 3 2.66 (1.56–4.52) < 0.001 Ref: Reference . C. Hyperparameter Optimization of ML Algorithms and Comparative Analysis To complement the conventional statistical analysis, three supervised machine‑learning algorithms—Logistic Regression (LR), Random Forest (RF), and Extreme Gradient Boosting (XGBoost)—were trained to predict the risk of common mental health disorders (CMHD) in the study population. The final optimized hyperparameter configurations for each model are presented in Table 4 . Table 4 Optimized value of the Algorithms Algorithms Optimized values of Hyperparameters LR 'C': 0.001, 'penalty': 'l2', 'solver': 'lbfgs RF 'max_depth': 10, 'max_features': 'log2', 'min_samples_split': 10, 'n_estimators': 500 XGBoost 'colsample_bytree': 0.8, 'learning_rate': 0.01, 'max_depth': 3, 'n_estimators': 200, 'subsample': 1.0 Model performance was evaluated using Accuracy, Area Under the Receiver Operating Characteristic Curve (AUC), and F1‑score, with Precision and Recall additionally reported to characterize classification trade‑offs. The detailed comparative results of the optimized models are summarized in Table 5 . Table 5 Comparative performance metrics of optimized machine learning algorithms. Model Accuracy AUC Score F1-Score Precision Recall Logistic Regression 0.762 0.746 0.865 0.762 1 Random Forest 0.776 0.757 0.863 0.807 0.926 XGBoost 0.771 0.763 0.864 0.788 0.957 Overall, the three optimized models showed similar F1‑scores (0.863–0.865), indicating a comparable balance between precision and recall in identifying CMHD in this dataset. The Random Forest model achieved the highest accuracy (0.776) and the highest precision (0.807), suggesting fewer false‑positive classifications relative to the other approaches. In contrast, XGBoost yielded the highest AUC (0.763), indicating the strongest overall discriminative ability across decision thresholds. Logistic Regression demonstrated perfect recall (1.000), meaning all CMHD cases were identified by the model, but this came with lower precision (0.762), consistent with a larger number of false positives. Taken together, the findings indicate that tree‑based ensemble models (RF and XGBoost) provided slightly stronger discrimination and/or precision, whereas Logistic Regression maximized sensitivity in this sample. Although XGBoost showed the highest AUC, Random Forest demonstrated the best overall balance across accuracy, precision, recall, and F1-score. Therefore, Random Forest was selected as the most appropriate algorithm for predicting CMHD, as shown in Figs. 5 and 6 . D. Variable Significance and Feature Contribution using Random Forest Since the Random Forest model showed the strongest overall predictive performance, we examined which predictors contributed most to its classification of CMHD risk. Feature contribution was quantified using the model’s built-in tree-based feature. To improve interpretability, importance values were aggregated at the variable level as presented in Fig. 7 . As shown in Fig. 7 , house loss was by far the most influential predictor (importance = 0.126), indicating that housing disruption is the dominant signal used by Random Forest when distinguishing individuals with CMHD from those without. The next most influential predictors were displacement (0.114) and land loss (0.101), highlighting the importance of erosion-related asset loss and mobility as major stress pathways. Exposure to erosion also contributed meaningfully (0.102), suggesting that direct environmental risk remains informative even after accounting for downstream losses. Beyond these primary shock-related factors, several household and psychosocial conditions contributed additional predictive information, including household income (0.099), number of child (0.058), and cope with that situation (cope_withh) (0.066). Overall, Fig. 7 shows that Random Forest predictive power is driven mainly by displacement-related shocks—especially house loss—followed by land loss and displacement history, with exposure, number of child, coping, and income providing incremental contributions to CMHD risk classification. 4. Discussion This is a population based cross-sectional study, offer a representative estimate of the prevalence of CMHD and explores its association with individuals living in river erosion-prone areas. A comprehensive review of the literature found that no similar study has been conducted in the country. This study was designed to fill that gap by measuring the prevalence of CMHD and identifying the key risk factors contributing to these mental health issues in populations displaced or affected by riverbank erosion. Understanding these factors is crucial for developing targeted interventions and support programs to improve mental health outcomes in these vulnerable communities. However, our study found that people exposed to river erosion had significantly higher rates of CMHDs based on the DAS score compared to those not exposed ( Figure. 3 ). These results align with earlier studies [ 16 , 33 – 35 ]. Thomas et al. [ 35 ] and Arobi et al. [ 36 ] found that people exposed to river erosion had a higher risk of developing DASS disorders compared to those not exposed. The reason is that exposed individuals lost various assets such as houses, property, and cattle, and some turned to drugs to cope with stress, anxiety, and frustration. Additionally, river erosion exposure may have worsened existing psychological distress, sleep problems, physical complaints, and psychosocial and behavioral issues [ 6 ]. We also found that the rates of abnormal Common Mental Health Disorders differed significantly among the three districts studied. Based on DAS scores, people in Rajbari and Tangail districts had a much higher risk of abnormal Common Mental Health Disorders compared to those in Kushtia district. One possible explanation is that during the dry season, people affected by erosion in Tangail and Rajbari districts had access to riverbeds along the Padma and Brahmaputra rivers due to their previous land boundaries. We also found that females were more likely to have DAS disorders than males, which coincided with earlier studies [ 33 , 37 ]. According to DAS scores, Common Mental Health Disorders were significantly more common among illiterate people than those with education, consistent with findings from Altaf et al. and Weyerer et al. [ 33 , 38 ]. Older age groups, especially those aged 38–45 and above 45, had higher rates of abnormal Common Mental Health Disorders compared to younger people (≤ 37 years), a pattern supported by previous research [ 33 , 39 ]. In the study population, we found that the prevalence of abnormal CMHD increased as the number of children in a family grew. This may be because having more children adds to the economic burden caused by river erosion. Respondents with larger families had higher chances of developing CMHD compared to those with fewer children, as managing a big family under financial stress was especially difficult for affected individuals. This finding aligns with previous research [ 33 ]. Additionally, the odds of having abnormal CMHD were higher among landless participants living within 0.2 miles of the riverbank compared to those who owned land and lived farther than 0.2 miles from the river. Further, we use advanced machine learning (ML) techniques to greatly extend traditional statistical methods by capturing high-dimensional, non-linear interactions between socio-environmental stresses. With a peak accuracy of 0.776 and a precision of 0.807, the Random Forest model proved to be the best classifier, surpassing the baseline Logistic Regression (AUC = 0.746). This discrepancy in performance suggests that complicated convergences, which are frequently oversimplified by traditional linear models, impact CMHD risk in populations affected by disasters. The feature importance analysis revealed "Loss of House" to be the main factor influencing CMHD classification, with the highest importance scores in the algorithmic framework, in line with the statistical evaluation. This finding indicates that the primary cause of psychological discomfort is the sudden loss of shelter. Algorithmic paths also demonstrated that secondary factors increase this risk; as the Lollipop Chart (Fig. 7 ) illustrates, significant contributions from Land Ownership and Displacement together boost the model's predictive power. From the standpoint of public health, the Random Forest model's high Accuracy (0.776). These results show that ML-driven predictive modelling is an effective tool for policymakers, despite the drawbacks of cross-sectional data and the "black-box" character of ensemble models. After erosion incidents, it makes it easier to quickly identify high-risk homes, enabling more focused mental health interventions. Study limitations This study had several limitations that should be considered when interpreting the findings. First, it was a cross-sectional study done in only three river erosion-prone districts (Kushtia, Rajbari, and Tangail) with a relatively small sample size, which means the findings may not apply to the whole country. Another limitation is that the study focused only on three major mental health conditions—depression, anxiety, and stress—as indicators of Common Mental Health Disorders (CMHD). Other important mental health issues, like neurodevelopmental disorders, sleep disorders, schizophrenia, bipolar disorder, obsessive-compulsive disorder, panic disorder, mixed anxiety and depression, and social phobia, were not covered in this study. 5. Conclusions This study demonstrates that people exposed to riverbank erosion had worse mental health and were more likely to develop common mental health disorders (CMHD) than those who weren’t exposed. Using advanced machine learning modeling, risk factors were identified, with the sudden “Loss of House”, “Loss of land”, and subsequent “Displacement” emerging as the most critical predictors of psychosocial distress. Additionally, our study highlights that certain demographic groups are more vulnerable to mental health disparities, such as women, the elderly, people without agricultural land, and those with bigger families (more than four children). Property loss and economic instability combine to create high-risk pathways for CMHD, as demonstrated by the Random Forest model's superior performance (Accuracy = 0.776), which highlights the complex and non-linear nature of these socio-environmental stressors. Special attention should also be given to vulnerable groups, including women, the elderly, landless individuals, and those with larger families, to reduce mental health inequalities. Introducing low-interest loans in erosion-prone areas could further help recovery and improve people's ability to adapt. Declarations Ethical approval The study’s procedures were conducted in accordance with the protocol approved by the Ethical Review Committee of Islamic University, Kushtia-7003, Bangladesh (Reference No. FS/IU/2025/17). All participants were informed prior to data collection about the study's objectives. The questionnaire was translated into the native language (Bangla) and read out to the respondents. When it was decided to conduct an interview, the educated respondents gave their written consent, and the uneducated participants signed the consent form with their fingerprints. Before taking part in this study, each participant gave their informed consent. Consent for publication Not applicable. Competing interests The authors declare no competing interests. Funding The authors did not receive any specific funding for this study. Author Contribution Conceptualization: Md. Tahidur Rahman and Md Lotifur Rohman; Data curation and statistical analysis: Md. Tahidur Rahman, Md Lotifur Rohman, Md. Maniruzzaman and Md. Merajul Islam; Drafting the manuscript: Md. Tahidur Rahman, Md Lotifur Rohman, Shyam Sundar Sarkar and Md. Fahim; Review and editing: Md. Tahidur Rahman, Md Lotifur Rohman, Md. Maniruzzaman and Md. Merajul Islam; Contribution to discussion, interpretation, and finalization: Md. Tahidur Rahman, Md Lotifur Rohman, Shyam Sundar Sarkar and Md. Fahim; Supervision: Md. Tahidur Rahman, Md. Maniruzzaman and Md. Merajul Islam; Validation. All authors have read and approved the final version of the manuscript. Acknowledgement The authors would like to acknowledge to Dr. Md. Sazzed Hossain Zahid, Associate professor, Department of English, Islamic University Kushtia, Bangladesh, for his constructive feedback and assistance in improving the language of the manuscript. Data Availability The datasets used in the current study are available from the corresponding author upon reasonable request. References Lee EH, Moon SH, Cho MS, Park ES, Kim SY, Han JS, Cheio JH. The 21-Item and 12-Item Versions of the Depression Anxiety Stress Scales: Psychometric Evaluation in a Korean Population. Asian Nurs Res (Korean Soc Nurs Sci). 2019;13. 10.1016/j.anr.2018.11.006 . Rapaport MH, Clary C, Fayyad R, Endicott J. Quality-of-life impairment in depressive and anxiety disorders. Am J Psychiatry. 2005;162. 10.1176/appi.ajp.162.6.1171 . Lovibond SH, Lovibond PF. Manual for the Depression Anxiety Stress Scales. 1995. WHO. Depressive disorder (depression). 2023. Baxter AJ, Vos T, Scott KM, Norman RE, Flaxman AD, Blore J, Whiteford HA. The regional distribution of anxiety disorders: Implications for the Global Burden of Disease Study, 2010. Int J Methods Psychiatr Res. 2014;23. 10.1002/mpr.1444 . Norris FH, Friedman MJ, Watson PJ, Byrne CM, Diaz E, Kaniasty K. 60,000 Disaster victims speak: Part I. An empirical review of the empirical literature, 1981–2001. Psychiatry. 2002;65. 10.1521/psyc.65.3.207.20173 . Costello A, Abbas M, Allen A, et al. Managing the health effects of climate change. Lancet and University College London Institute for Global Health Commission. Lancet. 2009;373. 10.1016/S0140-6736(09)60935-1 . Watts N, Amann M, Ayeb-Karlsson S, et al. The Lancet Countdown on health and climate change: from 25 years of inaction to a global transformation for public health. Lancet. 2018;391. 10.1016/S0140-6736(17)32464-9 . Ogbo FA, Mathsyaraja S, Koti RK, Perz J, Page A. The burden of depressive disorders in South Asia, 1990–2016: findings from the global burden of disease study. BMC Psychiatry. 2018;18. 10.1186/s12888-018-1918-1 . WHO: National Mental Health Survey of Bangladesh, 2018–2019: provisional fact sheet. Dhaka. 2019. Published Online First: 2019. Masum JH. Climatic Hazards in Bangladesh-A Literature Review Jahangir Hasan Masum. Published Online First. 2020. 10.13140/RG.2.2.10239.94882 . Munna TI. River erosion: A silent disaster in Bangladesh. 2019. Kamal M. Riverbank Erosion and Migration: A Study on Displaced People from Governance and Cultural Perspective. 2019, XIII:23–34. Khatun F, Bhattacharya D, Rahman M, Moazzem KG, Khan TI, Sabbih MA, Saadat S. Four Years of SDGs in Bangladesh: Measuring progress and charting the path forward. 2020. IPCC: Intergovernmental Panel on Climate Change working group II. Climate Change 2014: Impacts, Adaptation, and Vulnerability. Part B: Regional Aspects, Polar regions. New York: Cambridge Univ; 2014. Nagandla K, Nalliah S, Yin L, et al. Prevalence and associated risk factors of depression, anxiety and stress in pregnancy. Int J Reprod Contracept Obstet Gynecol Published Online First. 201610.18203/2320 – 1770.ijrcog20162132. Meltzer H, Bebbington P, Brugha T, Farrell M, Jenkins R. The relationship between personal debt and specific common mental disorders. Eur J Public Health. 2013;23:108–13. 10.1093/eurpub/cks021 . Darwish MA, Al Turki ZA, Sabra AA. Relation between Common Mental Disorders and Body Weight using Arabic DASS 21 scale (Depression, Anxiety and Stress Scale) among Adult Saudi Females Attending Primary Care. East Saudi Arabia. 2014. Sekoni O, Mall S, Christofides N. The relationship between protective factors and common mental disorders among female urban slum dwellers in Ibadan, Nigeria. PLoS ONE. 2022;17:1–16. 10.1371/journal.pone.0263703 . NIH. Depression: Transforming the understanding and treatment of mental illnesses. 2024. Park SG, Min KB, Chang SJ, Kim HC, Min JY. Job stress and depressive symptoms among Korean employees: The effects of culture on work. Int Arch Occup Environ Health. 2009;82. 10.1007/s00420-008-0347-8 . Eske J. What is acute stress disorder? MedicalNewsToday. 2019. Pearson K. X. On the criterion that a given system of deviations from the probable in the case of a correlated system of variables is such that it can be reasonably supposed to have arisen from random sampling. London, Edinburgh, Dublin Philos Mag J Sci. 1900, 50:157–75. Lavalley MP. Statistical Primer for Cardiovascular Research Logistic Regression. 2008, 2395–9. 10.1161/CIRCULATIONAHA.106.682658 Desarrollo U, Desarrollo U, Desarrollo U. Integration of RNA Editing with Multiomics Data Improves Machine Learning Models for Predicting Drug Responses in Breast Cancer Patients. 2024. Chen T, Guestrin C. XGBoost: A scalable tree boosting system. Proc ACM SIGKDD Int Conf Knowl Discov Data Min. 2016, 13-17-Augu:785–94. 10.1145/2939672.2939785 Liashchynskyi P. Grid Search, Random Search, Genetic Algorithm: A Big Comparison for NAS. arXiv Prepr ArXiv191206059. 2019. Yang L, Shami A. On Hyperparameter Optimization of Machine Learning Algorithms: Theory and Practice. 1–69. Swets JA, Swets JA. Measuring the Accuracy of Diagnostic Systems Linked references are available on JSTOR for this article: Measuring the Accuracy of Diagnostic Systems. 1988, 240:1285–93. Lobo JM, Jiménez-valverde A, Real R. AUC: a misleading measure of the performance of predictive distribution models. 2007, 1–7. 10.1111/j.1466-8238.2007.00358.x Sokolova M. Beyond Accuracy, F-score and ROC : a Family of Discriminant Measures for Performance Evaluation. Buckland M, Gey F. The Relationship between Recall and Precision. Published Online First: 1994. 10.1002/(sici)1097–4571(199401)45. Hossain A, Alam MJ, Haque MR. Effects of riverbank erosion on mental health of the affected people in Bangladesh. PLoS ONE. 2021;16. 10.1371/journal.pone.0254782 . Munro A, Kovats RS, Rubin GJ, et al. Effect of evacuation and displacement on the association between flooding and mental health outcomes: a cross-sectional analysis of UK survey data. Lancet Planet Heal. 2017;1. 10.1016/S2542-5196(17)30047-5 . Jom Thomas J, Prakash B, Kulkarni P, Narayana Murthy MR. Prevalence and severity of depression among people residing in flood affected areas of Kerala. Int J Community Med Public Heal. 2019;6. 10.18203/2394–6040.ijcmph20190600 . SarzamArobi, Naher J, Soron TR. Impact of River Bank Erosion on Mental Health and Coping Capacity in Bangladesh. Glob Psychiatry. 2019;2. 10.2478/gp-2019-0011 . Yeshaw Y, Mossie A. Depression, anxiety, stress, and their associated factors among Jimma university staff, Jimma, southwest Ethiopia, 2016: A cross-sectional study. Neuropsychiatr Dis Treat. 2017;13. 10.2147/NDT.S150444 . Weyerer S, Eifflaender-Gorfer S, Köhler L, et al. Prevalence and risk factors for depression in non-demented primary care attenders aged 75 years and older. J Affect Disord. 2008;111. 10.1016/j.jad.2008.02.008 . Pinheiro KAT, Horta BL, Pinheiro RT, Horta LL, Terres NG, Da Silva RA. Common mental disorders in adolescents: A population based cross-sectional study. Rev Bras Psiquiatr. 2007;29. 10.1590/s1516-44462006005000040 . Additional Declarations No competing interests reported. Cite Share Download PDF Status: Under Review Version 1 posted Reviewers agreed at journal 10 May, 2026 Reviewers invited by journal 29 Apr, 2026 Editor assigned by journal 28 Apr, 2026 Editor invited by journal 27 Apr, 2026 Submission checks completed at journal 25 Apr, 2026 First submitted to journal 25 Apr, 2026 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-9455372","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":634550257,"identity":"a7913006-ea8e-4cb2-ac8c-81767804d1bd","order_by":0,"name":"Md. Tahidur Rahman","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA1UlEQVRIiWNgGAWjYFACHoYDPECKjYH5AJCSkCFCA1wLWwJICw9RWhggyngMIAKEgD372YMH3rbdY+CTyPn86kaNBQ8D++GjG/DawpOXcHBuWzEDm0TuNuucY0CH8aSl3cDvsByDw7xtCWAtxjlsQC0SPGb4tfC/gWnJeWac848YLRJwW3KYH+e2EaPlxhuDg3POJfCw8TwzY87tkwAyCPiFvT/H+MObsgQ5+fbkx59zvtXJ8bMfPoZXC9w2IGaTALHYiFEOA8wfSFE9CkbBKBgFIwcAADvDPtIzFVNxAAAAAElFTkSuQmCC","orcid":"","institution":"Islamic University","correspondingAuthor":true,"prefix":"","firstName":"Md.","middleName":"Tahidur","lastName":"Rahman","suffix":""},{"id":634550258,"identity":"09eb00ab-fcc8-4b96-926f-d9765970572f","order_by":1,"name":"Md. Lotifur Rohman","email":"","orcid":"","institution":"Islamic University","correspondingAuthor":false,"prefix":"","firstName":"Md.","middleName":"Lotifur","lastName":"Rohman","suffix":""},{"id":634550259,"identity":"8cf1bdb3-9ced-4cc6-8b9d-77bd8ced0c5a","order_by":2,"name":"Shyam Sundar Sarkar","email":"","orcid":"","institution":"Islamic University","correspondingAuthor":false,"prefix":"","firstName":"Shyam","middleName":"Sundar","lastName":"Sarkar","suffix":""},{"id":634550262,"identity":"c38bcfff-6065-45a0-9b80-bfbd9d4b6a80","order_by":3,"name":"Md. Fahim","email":"","orcid":"","institution":"R.P. Shaha University","correspondingAuthor":false,"prefix":"","firstName":"Md.","middleName":"","lastName":"Fahim","suffix":""},{"id":634550263,"identity":"2d9a3a46-9ee2-43f2-98f2-f6c2fd1cffb3","order_by":4,"name":"Md. Merajul Islam","email":"","orcid":"","institution":"Jatiya Kabi Kazi Nazrul Islam University","correspondingAuthor":false,"prefix":"","firstName":"Md.","middleName":"Merajul","lastName":"Islam","suffix":""},{"id":634550265,"identity":"3cbebbb7-1641-455c-8411-62a18f73fce5","order_by":5,"name":"Md. Maniruzzaman","email":"","orcid":"","institution":"Khulna University","correspondingAuthor":false,"prefix":"","firstName":"Md.","middleName":"","lastName":"Maniruzzaman","suffix":""}],"badges":[],"createdAt":"2026-04-18 08:08:35","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-9455372/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-9455372/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":108780118,"identity":"bd1ea720-d86c-4834-b67b-a3d9a20e873d","added_by":"auto","created_at":"2026-05-08 10:06:40","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":219264,"visible":true,"origin":"","legend":"\u003cp\u003eDistribution of the sampled households.\u003c/p\u003e","description":"","filename":"floatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-9455372/v1/6e0063f9f5168d63ae7a3759.png"},{"id":108806936,"identity":"1f5db5c8-d04e-45dd-85f0-69f8d05875c4","added_by":"auto","created_at":"2026-05-08 15:29:41","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":234417,"visible":true,"origin":"","legend":"\u003cp\u003eA proposed ML-based framework for the prediction of CMHD.\u003c/p\u003e","description":"","filename":"floatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-9455372/v1/751303e5971c26d4f902dd44.png"},{"id":108780120,"identity":"0eae4ef9-0dc9-4958-9d5f-9634fa4e3685","added_by":"auto","created_at":"2026-05-08 10:06:40","extension":"jpg","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":40778,"visible":true,"origin":"","legend":"\u003cp\u003eProportion of respondents with Common Mental Health Disorders in exposed and non-exposed groups: N =497.\u003c/p\u003e","description":"","filename":"3.jpg","url":"https://assets-eu.researchsquare.com/files/rs-9455372/v1/50f79d405e0c8f2848fc57a0.jpg"},{"id":108807085,"identity":"99a47aab-846a-4c62-9937-d4a666f837c6","added_by":"auto","created_at":"2026-05-08 15:30:07","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":212086,"visible":true,"origin":"","legend":"\u003cp\u003eFlow of socio- environmental factors toward CMHD Status\u003c/p\u003e","description":"","filename":"floatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-9455372/v1/c08daf4c8a449ff66bf15592.png"},{"id":108780122,"identity":"50878bf1-f062-4744-8a58-bbc08435d276","added_by":"auto","created_at":"2026-05-08 10:06:40","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":66708,"visible":true,"origin":"","legend":"\u003cp\u003eComparison of Receiver Operating Characteristic (ROC) curves for three machine learning models in predicting CMHD\u003c/p\u003e","description":"","filename":"floatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-9455372/v1/987c532ae19b59b677e591f3.png"},{"id":108780123,"identity":"5955622f-5375-4269-a2eb-92c2174abff9","added_by":"auto","created_at":"2026-05-08 10:06:40","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":51649,"visible":true,"origin":"","legend":"\u003cp\u003eComparison of Tuned Model Performance Metrics for CMHD Prediction\u003c/p\u003e","description":"","filename":"floatimage5.png","url":"https://assets-eu.researchsquare.com/files/rs-9455372/v1/9652e1ba28bc169956d4896f.png"},{"id":108780124,"identity":"9747cda1-bea3-4491-af40-d4d457872657","added_by":"auto","created_at":"2026-05-08 10:06:40","extension":"png","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":43022,"visible":true,"origin":"","legend":"\u003cp\u003eCMHD Features Contribution (Random Forest).\u003c/p\u003e","description":"","filename":"floatimage6.png","url":"https://assets-eu.researchsquare.com/files/rs-9455372/v1/8010cae9b401b35d94cfc730.png"},{"id":109067672,"identity":"61c80027-0343-413f-a462-1240fa11b4d3","added_by":"auto","created_at":"2026-05-12 09:59:34","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1210298,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-9455372/v1/bb3dc767-a919-4c00-b743-f703b00077e2.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Identifying Associated Factors of Common Mental Health Disorders in Riverbank Erosion Areas of Bangladesh Using Machine Learning Algorithms","fulltext":[{"header":"1. Introduction","content":"\u003cp\u003eAccording to the World Health Organization (WHO), depression and anxiety are the two most Common Mental Health Disorders globally, affecting approximately 300\u0026nbsp;million people worldwide [\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e]. These disorders can impair daily functioning, lower work or academic performance, strain relationships, and, in severe cases, lead to suicide. At the societal level, they contribute to a significant economic burden [\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e]. The Depression Anxiety Stress Scales (DASS) were systematically developed based on the tripartite model of depression and anxiety to assess symptoms of Common Mental Health Disorders (CMHDs) [\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e]. Globally, it is estimated that about 280\u0026nbsp;million people suffer from depression, while around 272.5\u0026nbsp;million are affected by anxiety disorders [\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e, \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eNatural disasters affect millions of people each year, not only physically but also psychologically [\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e]. Among such disasters, riverbank erosion has been associated with increased rates of depression, anxiety, and stress in both developed and developing countries. Climate change has further intensified the frequency and severity of such extreme weather events, including riverbank erosion, which now occur beyond the range of normal natural variability, thereby disrupting social, economic, and ecological systems [\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e, \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e]. In South Asia, Bangladesh reports a higher prevalence of mental illness compared to neighbouring countries [\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e]. Among Bangladeshi adults, 16.8% of males and 17% of females experience mental health issues, yet 32.3% of them do not seek medical care [\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e]. Around 20 of the 64 districts in Bangladesh are highly vulnerable to riverbank erosion [\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e]. In 2019 alone, approximately 5,081 households were displaced, and nearly 2,270 hectares of land were lost due to erosion [\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e, \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e]. According to Khatun et al. [\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e], if natural disasters keep happening more often and become stronger, about 30,366,230 households in Bangladesh could be affected by 2030. These extreme weather events contribute not only to widespread displacement and forced migration but also to deteriorating conditions that influence both physical and mental well-being [\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e]. We conducted a comprehensive literature review and found no population-based study in Bangladesh that specifically investigates the prevalence and risk factors of CMHDs in river-prone areas. Globally, only a limited number of studies have explored the common mental health impacts in other populations, such as the prevalence and predictors of CMHDs among pregnant women in Malaysia and a rural community-based cohort of northern Vietnamese women [\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e, \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e] and the association between Common Mental Health Disorders and body weight in Eastern Saudi Arabia [\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e]. To the best of our knowledge, no prior research in Bangladesh or elsewhere has identified the key risk factors of Common Mental Health Disorders (CMHDs), based on Depression, Anxiety, and Stress (DAS) scores, among individuals affected by riverbank erosion. Therefore, this study aims to investigate how river erosion influences Common Mental Health Disorders in Bangladesh and to identify the risk factors associated with these conditions among affected populations.\u003c/p\u003e"},{"header":"2. Methods","content":"\u003cp\u003e \u003cb\u003eA. Study area and sampling procedure\u003c/b\u003e \u003c/p\u003e \u003cp\u003eWe purposefully chose three districts of Kushtia, Tangail, and Rajbari due to the high risk of river erosion. We initially selected three erosion-prone Upazila Kumarkhali in the Kushtia district, Bhuyapur Upazila in the Tangail district, and Rajbari Sadar Upazila in the Rajbari district. From each of these subdistricts, one union was selected: Shiladah Union from Kumarkhali, Arjuna Union from Bhuyapur, and Mijanpur Union from Rajbari Sadar. We then randomly selected seven villages (Ramkrisnopur, Charjoukuri, Ramcondropur, Chilimpur, Dhunchi, Sonakandor and Charnarayanpur) from Shiladah union, two villages (Kalua, Kumarkandi) from Arjuna union, and four villages (Arjuna, Tarai, Dhubliya, and Kuthiboira) from Mijanpur union. A total of 13 villages were chosen from the three districts. The detail description of the selected household sample is presented in Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e. We calculated the sample size using the formula for a two-stage cluster sampling design:\u003c/p\u003e \u003cp\u003e \u003cb\u003eStep 1: Calculate initial sample size (\u003c/b\u003e \u003cspan class=\"InlineEquation\"\u003e \u003cspan class=\"mathinline\"\u003e\\(\\:{n}_{0}\\)\u003c/span\u003e \u003c/span\u003e \u003cb\u003e)\u003c/b\u003e:\u003cdiv id=\"Equa\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equa\" name=\"EquationSource\"\u003e\n$$\\:{n}_{0}=\\frac{{Z}^{2}P(1-P)}{{d}^{2}}$$\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e \u003cp\u003eWhere:\u003c/p\u003e \u003cp\u003e \u003cul\u003e \u003cli\u003e \u003cp\u003e \u003cspan class=\"InlineEquation\"\u003e \u003cspan class=\"mathinline\"\u003e\\(\\:Z=1.96\\)\u003c/span\u003e \u003c/span\u003e (for 95% confidence), \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:P=0.5\\)\u003c/span\u003e\u003c/span\u003e (assumed prevalence), \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:d=0.05\\)\u003c/span\u003e\u003c/span\u003e (margin of error) [\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e]\u003c/p\u003e \u003c/li\u003e \u003c/ul\u003e \u003cdiv id=\"Equb\" class=\"Equation\"\u003e \u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equb\" name=\"EquationSource\"\u003e\n$$\\:{n}_{0}=\\frac{(1.96{)}^{2}\\times\\:0.5\\times\\:0.5}{(0.05{)}^{2}}=\\frac{3.8416\\times\\:0.25}{0.0025}=\\frac{0.9604}{0.0025}=384.16\\approx\\:384$$\u003c/div\u003e \u003c/div\u003e \u003c/p\u003e \u003cp\u003eSo, the initial sample size \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{n}_{0}=384\\)\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e \u003cp\u003e \u003cb\u003eStep 2: Adjust for design effect (\u003c/b\u003e \u003cspan class=\"InlineEquation\"\u003e \u003cspan class=\"mathinline\"\u003e\\(\\:Deff\\)\u003c/span\u003e \u003c/span\u003e \u003cb\u003e)\u003c/b\u003e:\u003cdiv id=\"Equc\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equc\" name=\"EquationSource\"\u003e\n$$\\:Deff=1+(m-1)\\rho\\:$$\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e \u003cp\u003eWhere:\u003c/p\u003e \u003cp\u003e \u003cul\u003e \u003cli\u003e \u003cp\u003e \u003cspan class=\"InlineEquation\"\u003e \u003cspan class=\"mathinline\"\u003e\\(\\:m=50\\)\u003c/span\u003e \u003c/span\u003e (average cluster size), \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\rho\\:=0.01\\)\u003c/span\u003e\u003c/span\u003e (intra-cluster correlation)\u003c/p\u003e \u003c/li\u003e \u003c/ul\u003e \u003cdiv id=\"Equd\" class=\"Equation\"\u003e \u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equd\" name=\"EquationSource\"\u003e\n$$\\:Deff=1+(50-1)\\times\\:0.01=1+49\\times\\:0.01=1+0.49=1.49\\approx\\:1.50$$\u003c/div\u003e \u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003cb\u003eStep 3: Apply design effect to get adjusted sample size (\u003c/b\u003e \u003cspan class=\"InlineEquation\"\u003e \u003cspan class=\"mathinline\"\u003e\\(\\:{n}_{\\text{deff}}\\)\u003c/span\u003e \u003c/span\u003e \u003cb\u003e)\u003c/b\u003e \u003cdiv id=\"Eque\" class=\"Equation\"\u003e \u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Eque\" name=\"EquationSource\"\u003e\n$$\\:{n}_{\\text{deff}}={n}_{0}\\times\\:Deff=384\\times\\:1.50=576$$\u003c/div\u003e \u003c/div\u003e \u003c/p\u003e \u003cp\u003eSo, the sample size after adjusting for the design effect is 576.\u003c/p\u003e \u003cp\u003e \u003cb\u003eStep 4: Adjust for non-response rate (\u003c/b\u003e \u003cspan class=\"InlineEquation\"\u003e \u003cspan class=\"mathinline\"\u003e\\(\\:r\\)\u003c/span\u003e \u003c/span\u003e \u003cb\u003e)\u003c/b\u003e \u003cdiv id=\"Equf\" class=\"Equation\"\u003e \u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equf\" name=\"EquationSource\"\u003e\n$$\\:{n}_{\\text{final}}={n}_{\\text{deff}}\\times\\:(1+r)=576\\times\\:(1+0.10)=576\\times\\:1.10=633.6\\approx\\:634$$\u003c/div\u003e \u003c/div\u003e \u003c/p\u003e \u003cp\u003eTherefore, the minimum required sample size was 634, and a total of 711 households were finally selected for the study to ensure greater precision, comprising 100 from Shiladha Union (Kushtia District), 201 from Arjuna Union (Tangail District), and 410 from Mijanpur Union (Rajbari district).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003cb\u003eB. Study variables\u003c/b\u003e \u003c/p\u003e \u003cp\u003e \u003cb\u003eOutcome variables\u003c/b\u003e: The outcome variable in this study is Common Mental Health Disorders (CMHD), defined based on three mental health indicators: depression, anxiety, and stress. These were assessed using the Depression Anxiety Stress Scales\u0026ndash;21 items (DASS-21), which comprises three subscales\u0026mdash;each with 7 items\u0026mdash;corresponding to depression, anxiety, and stress. Participants responded to each item on a 4-point Likert scale ranging from \u003cb\u003e0\u003c/b\u003e (\"Did not apply to me at all\") to 3 (\"Applied to me very much or most of the time\"). The total score for each subscale ranges from 0 to 21 and \u003cb\u003eis\u003c/b\u003e multiplied by two (\u0026times;2) to align with the DASS scoring guidelines. The following cut-off scores were used to determine the presence of CMHD symptoms: Depression: \u0026ge;10, Anxiety: \u0026ge;8, Stress: \u0026ge;15 [\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e, \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e]. We defined a binary variable, where \"1\" indicates the presence of Common Mental Health Disorders (CMHD), and \"0\" indicates their absence.\u003c/p\u003e \u003cp\u003e \u003cb\u003eC. Measures\u003c/b\u003e \u003c/p\u003e \u003cp\u003e \u003cb\u003eDepression, Anxiety and Stress Scales\u003c/b\u003e \u003c/p\u003e \u003cp\u003e \u003cstrong\u003eDepression\u003c/strong\u003e \u003cp\u003eInvolves symptoms like sadness, loss of interest, sleep or appetite problems, guilt, trouble concentrating, and fatigue. It can be caused by genetic, biological, environmental, sociodemographic, and psychological factors [\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e].\u003c/p\u003e \u003c/p\u003e \u003cp\u003e \u003cstrong\u003eAnxiety\u003c/strong\u003e \u003cp\u003eAnxiety is a reaction to perceived threats, marked by troubling thoughts, tension, fast heartbeat, sweating, dizziness, and chest pain. It may result from natural disasters, poverty, family issues, personality traits, or brain chemistry [\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e].\u003c/p\u003e \u003c/p\u003e \u003cp\u003e \u003cstrong\u003eStress\u003c/strong\u003e \u003cp\u003earises when people feel overwhelmed by demands. Symptoms include mood changes, flashbacks, avoidance, and heightened alertness. Causes include trauma, loss, serious illness, or accidents [\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e].\u003c/p\u003e \u003c/p\u003e \u003cp\u003e \u003cb\u003ePredictor variables\u003c/b\u003e: The predictor variables included socioeconomic, demographic, and riverbank erosion-related factors, as well as the household's exposure status. Riverbank erosion-related variables included the distance of the homestead from the river (in miles), time since last displacement, amount of agricultural land (yes/no), loss of agricultural land (yes/no), cattle loss (yes/no), death of a relative (yes/no), home loss (yes/no), social detachment (yes/no), substance misuse (yes/no), social encouragement for rehabilitation (yes/no), and hope for land recovery (yes/no). Demographic characteristics comprised sex (male/female), age group (in years), and number of children. Socioeconomic factors included the respondent's occupation, household income, and educational status. Displacement, a direct consequence of riverbank erosion, was categorized into three groups based on the time since the last displacement: non-displaced, displaced within the past three years, and displaced more than three years ago. Age was categorized into three groups: 22\u0026ndash;37 years, 38\u0026ndash;45 years, and 46\u0026ndash;80 years. Similarly, the number of children was grouped into three categories: 1\u0026ndash;2, 3\u0026ndash;4, and more than 4.\u003c/p\u003e \u003cp\u003e \u003cb\u003eD. Feature Selection\u003c/b\u003e \u003c/p\u003e \u003cp\u003eFeature selection was performed to identify the most relevant predictors for the machine learning models. In the first stage, a chi-square test was applied to identify variables significantly associated with the outcome variable, and only statistically significant variables held. In the second stage, logistic regression analysis was conducted on those chi-square significant variables to further assess their independent effect. A p-value\u0026thinsp;\u0026lt;\u0026thinsp;0.05 was used to determine the statistical significance of these associations. These selected features were subsequently used for model development and evaluation on the testing dataset.. All statistical analyses were conducted using Stata version 15 (StataCorp, College Station, TX, USA) and Python (Jupyter Notebook; ipykernel 6.25.1).\u003c/p\u003e \u003cp\u003e \u003cb\u003eE. Machine Learning Framework and Algorithms\u003c/b\u003e \u003c/p\u003e \u003cp\u003eThe overview of the proposed machine-learning (ML) framework for predicting common mental health disorders (CMHD) is illustrated in Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e. First, the complete dataset (including participants with and without CMHD) is randomly divided into two subsets: 70% training data and 30% testing data. Using the training subset, a univariate feature screening step is performed in which each candidate predictor is evaluated against the CMHD outcome using the Chi-square test of independence, and p-values are computed for all features. Features meeting the predefined significance threshold (p\u0026thinsp;\u0026lt;\u0026thinsp;0.05) are retained as the optimal feature set [\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eNext, the reduced training data (containing only the optimal features) is used to build an ML‑based prediction system by training three classifiers: Logistic Regression (LR), Random Forest (RF), and XGBoost\u003c/p\u003e \u003cp\u003e \u003col style=\"list-style-type:lower-roman;\"\u003e \u003cspan\u003e \u003cli\u003e \u003cp\u003eLogistic Regression (LR): Utilized as a linear baseline to compare against ensemble methods [\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e].\u003c/p\u003e \u003c/li\u003e \u003c/span\u003e \u003cspan\u003e \u003cli\u003e \u003cp\u003eRandom Forest (RF): An ensemble bagging technique that builds several decision trees to manage feature interactions and increase classification accuracy [\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e].\u003c/p\u003e \u003c/li\u003e \u003c/span\u003e \u003cspan\u003e \u003cli\u003e \u003cp\u003eXGBoost: A gradient boosting framework that builds sequential trees to minimize residual errors through an iterative optimization [\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e].\u003c/p\u003e \u003c/li\u003e \u003c/span\u003e \u003c/ol\u003e \u003c/p\u003e \u003cp\u003e \u003cb\u003eF. Hyperparameter Tuning and Selection Criteria\u003c/b\u003e \u003c/p\u003e \u003cp\u003eGrid search [\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e] with fivefold cross-validation was used for systematic hyperparameter optimization to ensure the robustness and generalizability of the prediction models. By dividing the training data into five separate subsets and iteratively training the models on four folds while validating on the remaining one, this method made it possible to thoroughly evaluate different parameter combinations. This method guarantees that the chosen parameters function consistently across several data segments and reduces the possibility of overfitting [\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e]. The hyperparameters to be used in optimization are given in Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eHyperparameter optimization of algorithms using the grid search method\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"2\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAlgorithms\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eSearch range of each parameter\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eLR\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e'C': [0.001, 0.01, 0.1, 1, 10, 100], 'penalty': ['10', '12','13', '14'], 'solver': ['lbfgs', 'liblinear']\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRF\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e'n_estimators': [100, 200, 300, 400, 450, 500, 550, 600], 'max_depth': [None, 10, 20], 'min_samples_split': [\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e, \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e, \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e, \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e], 'max_features': ['sqrt', 'log2']\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eXGBoost\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e'n_estimators': [100, 200, 300, 400], 'max_depth': [\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e, \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e, \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e], 'learning_rate': [0.01, 0.1, 0.2], 'subsample': [0.8, 1.0], 'colsample_bytree': [0.8, 0.9, 1.0, 1.2]\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003eThe model parameters were estimated through maximum likelihood estimation (MLE) using the training dataset. The fitted model was subsequently employed to predict class membership (CMHD vs. healthy controls) in the test dataset. Furthermore, class probabilities for the response variable were computed to quantify the likelihood of each predicted outcome.\u003c/p\u003e \u003cp\u003e \u003cb\u003eG. Model Evaluation and Feature Importance\u003c/b\u003e \u003c/p\u003e \u003cp\u003eA wide range of evaluation metrics were used to evaluate the optimized models' performance:\u003c/p\u003e \u003cp\u003e \u003col style=\"list-style-type:lower-roman;\"\u003e \u003cspan\u003e \u003cli\u003e \u003cp\u003eAccuracy: The model's overall accuracy [\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e].\u003c/p\u003e \u003c/li\u003e \u003c/span\u003e \u003cspan\u003e \u003cli\u003e \u003cp\u003eThe model's capacity to distinguish between CMHD and non-CMHD situations is measured by the Area under the Receiver Operating Characteristic Curve (AUC) [\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e].\u003c/p\u003e \u003c/li\u003e \u003c/span\u003e \u003cspan\u003e \u003cli\u003e \u003cp\u003eF1-Score: The harmonic mean of recall and precision, guaranteeing a fair assessment for classification [\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e].\u003c/p\u003e \u003c/li\u003e \u003c/span\u003e \u003cspan\u003e \u003cli\u003e \u003cp\u003eRecall: Particularly given top priority in order to identify as many susceptible people as possible [\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e].\u003c/p\u003e \u003c/li\u003e \u003c/span\u003e \u003c/ol\u003e \u003c/p\u003e"},{"header":"3. Results","content":"\u003cp\u003eData were collected from 711 households to assess common mental health conditions in river erosion\u0026ndash;affected areas of Bangladesh. The dataset was randomly divided into a training set (70%; n\u0026thinsp;=\u0026thinsp;497) and a testing set (30%; n\u0026thinsp;=\u0026thinsp;214). All subsequent analyses were conducted using the training sample (n\u0026thinsp;=\u0026thinsp;497).\u003c/p\u003e \u003cp\u003e \u003cb\u003eA. Bivariate analysis of CMHD and associated factors\u003c/b\u003e \u003c/p\u003e \u003cp\u003eThe largest proportion of participants was from the Rajbari district (n\u0026thinsp;=\u0026thinsp;281; 56.54%), followed by Tangail (n\u0026thinsp;=\u0026thinsp;141; 28.37%) and Kushtia (n\u0026thinsp;=\u0026thinsp;75; 15.09%). Approximately 47.48% of respondents were uneducated, and 78.27% did not own any agricultural land. Of the total respondents, 397 (80.08%) were exposed to riverbank erosion, while 99 (19.92%) were not. Among the total participants, 6.64% reported the loss of cattle, 10.87% lost a relative, 44.27% lost their homestead, and 77.06% lost agricultural land due to erosion (Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e). According to regional distribution, Rajbari had the highest percentage of respondents (56.54%), followed by Tangail (28.37%) and Kushtia (15.09%) (Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eFigure \u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e illustrates that among the exposed respondents, the highest proportion (84.17%) exhibited abnormal mental health conditions, indicating a strong association between erosion exposure and mental disorders.\u003c/p\u003e \u003cp\u003eThe bivariate distribution of Common Mental Health Disorders (CMHD) across different predictors is presented in Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e. We observed that the proportion of CMHD was higher in Rajbari (80.43%) and Tangail (78.01%) than in Kushtia (57.33%), a difference that is statistically significant (p\u0026thinsp;\u0026lt;\u0026thinsp;0.001). The prevalence of CMHD increased with age, peaking at 84.66% in the oldest age group (46\u0026ndash;80 years; p\u0026thinsp;=\u0026thinsp;0.007). Compared to their counterparts, respondents who were uneducated (80.83%), had more than four children (90.20%), did not own agricultural land (80.46%), and had lower monthly earnings (\u0026thinsp;\u0026gt;\u0026thinsp;=\u0026thinsp;6000\u003cspan\u003e$\u003c/span\u003e taka; 82.93%) showed the highest rates of CMHD. These differences were statistically significant (p\u0026thinsp;\u0026lt;\u0026thinsp;0.05).By profession, the highest rates of CMHD were found among day laborers (87.23%) and drivers (83.33%). Among respondents who were exposed and relocated within the last three years, the prevalence of CMHD was significantly higher (97.56%) compared to those who did not migrate (59.01%; p\u0026thinsp;\u0026lt;\u0026thinsp;0.001). We found that the prevalence of CMHD was statistically higher among those who lost their cattle (93.94% vs. 75.00%, p\u0026thinsp;=\u0026thinsp;0.023), houses (94.55% vs. 61.73%, p\u0026thinsp;\u0026lt;\u0026thinsp;0.001), and agricultural land (84.86% vs. 47.37%, p\u0026thinsp;\u0026lt;\u0026thinsp;0.001) than among those who did not experience such losses. Furthermore, respondents who required social encouragement (88.56% vs. 67.91%, p\u0026thinsp;\u0026lt;\u0026thinsp;0.001), experienced social detachment (91.82% vs. 68.93%, p\u0026thinsp;\u0026lt;\u0026thinsp;0.001), and had hope for land return (87.25% vs. 73.42%, p\u0026thinsp;\u0026lt;\u0026thinsp;0.001) also exhibited significantly higher rates of CMHD.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab2\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003ePrevalence of common mental health disorders with different predictors.\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"6\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003eCharacteristics\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\" morerows=\"2\" rowspan=\"3\"\u003e \u003cp\u003eOverall\u003c/p\u003e \u003cp\u003eN\u0026thinsp;=\u0026thinsp;497(100%) n (%)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colspan=\"2\" nameend=\"c4\" namest=\"c3\"\u003e \u003cp\u003eCommon mental health disorders\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003e\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\chi\\:}^{2}\\)\u003c/span\u003e\u003c/span\u003e-value\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\" morerows=\"2\" rowspan=\"3\"\u003e \u003cp\u003ep-value\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eNormal\u003c/p\u003e \u003cp\u003eN\u0026thinsp;=\u0026thinsp;118 (23.74%)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eAbnormal\u003c/p\u003e \u003cp\u003eN\u0026thinsp;=\u0026thinsp;379 (76.26%)\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003en (%)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003en (%)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"6\" nameend=\"c6\" namest=\"c1\"\u003e \u003cp\u003e\u003cb\u003eSocio-demographic and economic factors\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eExposed\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"5\" nameend=\"c6\" namest=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eYes\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e398 (80.08)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e63 (15.83)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e335 (84.17)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003e66.929\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNo\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e99 (19.92)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e55 (55.56)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e44 (44.44)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eArea\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"5\" nameend=\"c6\" namest=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eKushtia\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e75 (15.09)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e32 (42.67)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e43 (57.33)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\" morerows=\"2\" rowspan=\"3\"\u003e \u003cp\u003e17.773\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\" morerows=\"2\" rowspan=\"3\"\u003e \u003cp\u003e\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\u0026lt;0.001\\)\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTangail\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e141 (28.37)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e31 (21.99)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e110 (78.01)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRajbari\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e281 (56.54)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e55 (19.57)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e226 (80.43)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"6\" nameend=\"c6\" namest=\"c1\"\u003e \u003cp\u003e\u003cb\u003eAge group(year)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e22\u0026ndash;37\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e170 (34.21)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e45 (26.47)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e125 (73.53)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\" morerows=\"2\" rowspan=\"3\"\u003e \u003cp\u003e9.825\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\" morerows=\"2\" rowspan=\"3\"\u003e \u003cp\u003e0.007\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e38\u0026ndash;45\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e164 (33.00)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e48 (29.27)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e116 (70.73)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e46\u0026ndash;80\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e163 (32.8)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e25 (15.34)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e138 (84.66)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"6\" nameend=\"c6\" namest=\"c1\"\u003e \u003cp\u003e\u003cb\u003eEducational status\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eUneducated\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e236 (47.48)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e45 (19.07)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e191 (80.83)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003e4.943\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003e0.026\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eEducated\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e261 (52.52)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e73 (27.97)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e188 (72.03)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"6\" nameend=\"c6\" namest=\"c1\"\u003e \u003cp\u003e\u003cb\u003eProfession\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eHousewife\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e230 (46.28)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e59 (25.65)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e171 (74.35)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\" morerows=\"6\" rowspan=\"7\"\u003e \u003cp\u003e9.709\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\" morerows=\"6\" rowspan=\"7\"\u003e \u003cp\u003e0.137\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eFarmer\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e62 (12.47)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e13 (20.97)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e49 (79.03)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eBusiness\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e55 (11.07)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e19 (34.55)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e36 (65.45)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDay labor\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e47 (9.46)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e6 (12.77)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e41 (87.23)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eService\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e23 (4.63)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e7 (30.43)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e16 (69.57)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDriver\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e30 (6.04)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e5 (16.67)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e25 (83.33)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eOthers\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e50 (10.06)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e9 (18.00)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e41 (82.00)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"6\" nameend=\"c6\" namest=\"c1\"\u003e \u003cp\u003e\u003cb\u003eNo of children\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e1\u0026ndash;2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e196 (39.44)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e52 (26.53)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e144 (73.47)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\" morerows=\"2\" rowspan=\"3\"\u003e \u003cp\u003e6.374\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\" morerows=\"2\" rowspan=\"3\"\u003e \u003cp\u003e0.041\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e3\u0026ndash;4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e250 (50.30)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e61 (24.40)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e189 (75.60)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u0026gt;\u0026thinsp;4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e51 (10.26)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e5 (9.80)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e46 (90.20)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"6\" nameend=\"c6\" namest=\"c1\"\u003e \u003cp\u003e\u003cb\u003eMonthly earnings (taka)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u0026le;\u0026thinsp;6000\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e41 (8.25)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e7 (17.07)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e34 (82.93)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\" morerows=\"3\" rowspan=\"4\"\u003e \u003cp\u003e13.752\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\" morerows=\"3\" rowspan=\"4\"\u003e \u003cp\u003e0.003\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e6001\u0026ndash;10000\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e157 (31.59)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e26 (16.56)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e131 (75.25)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e10001\u0026ndash;15000\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e202 (40.64)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e50 (24.75)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e152 (83.44)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u0026ge;\u0026thinsp;15000\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e97 (19.52)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e35 (36.08)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e62 (63.92)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"6\" nameend=\"c6\" namest=\"c1\"\u003e \u003cp\u003e\u003cb\u003eAgricultural land\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNo\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e389 (78.27)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e76 (19.54)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e313 (80.46)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003e16.432\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eYes\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e108 (21.73)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e42 (38.89)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e66 (61.11)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"6\" nameend=\"c6\" namest=\"c1\"\u003e \u003cp\u003e\u003cb\u003eAmount of agricultural land (decimal)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNo agricultural land\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e389 (78.27)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e76 (19.54)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e313 (80.46)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\" morerows=\"3\" rowspan=\"4\"\u003e \u003cp\u003e18.43\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\" morerows=\"3\" rowspan=\"4\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e1\u0026ndash;66\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e77 (15.49)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e28 (36.36)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e49 (63.64)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e67\u0026ndash;99\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e11 (2.21)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e5 (45.45)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e6 (54.55)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u0026ge;\u0026thinsp;100\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e20 (4.02)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e9 (45.00)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e11 (55.0)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"6\" nameend=\"c6\" namest=\"c1\"\u003e \u003cp\u003e\u003cb\u003eFactors contributing to riverbank erosion\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"6\" nameend=\"c6\" namest=\"c1\"\u003e \u003cp\u003e\u003cb\u003eExposure category\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eExposed\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e398 (80.08)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e63 (15.83)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e335 (84.17)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003e66.929\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNon-exposed\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e99 (19.92)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e55 (55.56)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e44 (44.44)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"6\" nameend=\"c6\" namest=\"c1\"\u003e \u003cp\u003e\u003cb\u003eDuration of internal displacement (Years)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNot migrate\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e222 (44.67)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e91 (40.99)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e131 (59.01)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\" morerows=\"2\" rowspan=\"3\"\u003e \u003cp\u003e78.148\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\" morerows=\"2\" rowspan=\"3\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u0026le;\u0026thinsp;3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e164 (33)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e4 (2.44)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e160 (97.56)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u0026gt;\u0026thinsp;3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e111 (22.33)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e23 (20.72)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e88 (79.28)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"6\" nameend=\"c6\" namest=\"c1\"\u003e \u003cp\u003e\u003cb\u003eDistance of homestead from the river (in miles)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u0026le;\u0026thinsp;0.2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\" morerows=\"1\" rowspan=\"2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003e0.978\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u0026gt;\u0026thinsp;0.2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"6\" nameend=\"c6\" namest=\"c1\"\u003e \u003cp\u003e\u003cb\u003eLoss of cattle\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNo\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e464 (93.36)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e116 (25.00)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e384 (75)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003e5.103\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003e0.023\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eYes\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e33 (6.64)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e2 (6.06)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e31 (93.94)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"6\" nameend=\"c6\" namest=\"c1\"\u003e \u003cp\u003e\u003cb\u003eDeath of relative\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNo\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e443 (89.13)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e333 (75.17)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e110 (24.83)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\" morerows=\"1\" rowspan=\"2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003e0.143\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eYes\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e54 (10.87)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e46 (85.19)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e8 (14.81)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"6\" nameend=\"c6\" namest=\"c1\"\u003e \u003cp\u003e\u003cb\u003eLoss of agricultural land\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNo\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e144 (22.94)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e60 (52.63)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e54 (47.37)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003e66.136\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eYes\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e383 (77.06)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e58 (15.14)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e325 (84.86)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"6\" nameend=\"c6\" namest=\"c1\"\u003e \u003cp\u003e\u003cb\u003eSubstance misuse\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNo\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e481 (96.78)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e117 (24.32)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e364 (75.68)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003e1.885\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003e0.1698\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eYes\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e16 (3.22)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1 (6.25)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e15 (93.75)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"6\" nameend=\"c6\" namest=\"c1\"\u003e \u003cp\u003e\u003cb\u003eSocial encouragement\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNo\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e296 (59.56)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e95 (32.09)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e201 (67.91)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003e27.07\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eYes\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e201 (40.44)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e23 (11.44)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e178 (88.56)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"6\" nameend=\"c6\" namest=\"c1\"\u003e \u003cp\u003e\u003cb\u003eLoss of home\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNo\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e277 (55.73)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e106 (38.27)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e171 (61.73)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003e71.114\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eYes\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e220 (44.27)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e12 (5.45)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e208 (94.55)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"6\" nameend=\"c6\" namest=\"c1\"\u003e \u003cp\u003e\u003cb\u003eSocial detachment\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNo\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e338 (68.01)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e105 (31.07)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e233 (68.93)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003e30.038\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eYes\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e159 (31.99)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e13 (8.18)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e146 (91.82)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"6\" nameend=\"c6\" namest=\"c1\"\u003e \u003cp\u003e\u003cb\u003eHope for land return\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNo\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e395 (79.48)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e105 (26.58)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e290 (73.42)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003e7.826\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eYes\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e102 (20.52)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e13 (12.75)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e89 (87.25)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003cem\u003eP-values for each variable were computed using the\u003c/em\u003e \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\chi\\:}^{2}\\)\u003c/span\u003e\u003c/span\u003e \u003cem\u003etest statistic.\u003c/em\u003e\u003c/p\u003e \u003cp\u003eA Parcats Plot (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e) illustrates a comprehensive snapshot of the study and visualizes the interaction between the most significant determinants identified by the chi-square tests: Displacement, Exposure, Loss land, and Loss house. The concentration of CMHD was significantly higher among those who had frequent displacements (\u0026gt;\u0026thinsp;3 times), even if the percentage of people with lower displacement frequency (\u0026thinsp;\u0026lt;\u0026thinsp;=\u0026thinsp;3 times) was higher. Participants who lost their homes showed the lowest levels of psychological resilience (green routes), while those who did not lose any assets showed the highest levels. Critical differences in stressors at the community, household, and environmental levels were further highlighted by the much greater trajectories toward CMHD of those who experienced the junction of high exposure and dual land-house loss compared to those who did not.\u003c/p\u003e \u003cp\u003e \u003cb\u003eB. Binary Logistic Regression Analysis of CMHD and Associated Factors\u003c/b\u003e \u003c/p\u003e \u003cp\u003eThe logistic regression analysis's findings are shown in Table\u0026nbsp;\u003cspan refid=\"Tab3\" class=\"InternalRef\"\u003e3\u003c/span\u003e. Higher probabilities of aberrant outcomes were highly correlated with exposure (OR\u0026thinsp;=\u0026thinsp;6.65, 95% CI: 4.12\u0026ndash;10.73; p\u0026thinsp;\u0026lt;\u0026thinsp;0.001). The probabilities of aberrant outcomes were considerably greater for participants who lost agricultural land (OR\u0026thinsp;=\u0026thinsp;6.23, 95% CI: 3.92\u0026ndash;9.88; p\u0026thinsp;\u0026lt;\u0026thinsp;0.001), livestock (OR\u0026thinsp;=\u0026thinsp;5.17, 95% CI: 1.22\u0026ndash;21.92; p\u0026thinsp;=\u0026thinsp;0.026), and home (OR\u0026thinsp;=\u0026thinsp;10.74, 95% CI: 5.72\u0026ndash;20.18; p\u0026thinsp;\u0026lt;\u0026thinsp;0.001). Conversely, having formal education (OR\u0026thinsp;=\u0026thinsp;0.61, 95% CI: 0.40\u0026ndash;0.93; p\u0026thinsp;=\u0026thinsp;0.020), having a larger monthly income (\u0026ge;\u0026thinsp;15000 taka: OR\u0026thinsp;=\u0026thinsp;0.36, 95% CI: 0.15\u0026ndash;0.91; p\u0026thinsp;=\u0026thinsp;0.030), owning agricultural land (OR\u0026thinsp;=\u0026thinsp;0.38, 95% CI: 0.24\u0026ndash;0.61; p\u0026thinsp;\u0026lt;\u0026thinsp;0.001). In addition to psychosocial characteristics including social encouragement (OR\u0026thinsp;=\u0026thinsp;3.66, 95% CI: 2.22\u0026ndash;6.02; p\u0026thinsp;\u0026lt;\u0026thinsp;0.001) and social detachment (OR\u0026thinsp;=\u0026thinsp;5.06, 95% CI: 2.74\u0026ndash;9.34; p\u0026thinsp;\u0026lt;\u0026thinsp;0.001), respondents with more than four children had greater odds (OR\u0026thinsp;=\u0026thinsp;3.32, 95% CI: 1.25\u0026ndash;8.82; p\u0026thinsp;=\u0026thinsp;0.016). Geographically, respondents from Rajbari (OR\u0026thinsp;=\u0026thinsp;1.90, 95% CI: 1.11\u0026ndash;3.25; p\u0026thinsp;=\u0026thinsp;0.019) and Tangail (OR\u0026thinsp;=\u0026thinsp;2.48, 95% CI: 1.33\u0026ndash;4.62; p\u0026thinsp;=\u0026thinsp;0.004) had greater odds than those from Kushtia. Increased risks of aberrant outcomes were also substantially correlated with older age (46\u0026ndash;80 years: OR\u0026thinsp;=\u0026thinsp;1.99, 95% CI: 1.15\u0026ndash;3.43; p\u0026thinsp;=\u0026thinsp;0.014) and internal displacement (\u0026le;\u0026thinsp;3 years: OR\u0026thinsp;=\u0026thinsp;27.79, 95% CI: 9.94\u0026ndash;77.64; p\u0026thinsp;\u0026lt;\u0026thinsp;0.001; \u0026gt;3 years: OR\u0026thinsp;=\u0026thinsp;2.66, 95% CI: 1.56\u0026ndash;4.52; p\u0026thinsp;\u0026lt;\u0026thinsp;0.001).\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab3\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 3\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eRisk Factors Associated with Common Mental Health Disorders: Binary Logistic Regression Analysis.\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"3\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCharacteristic\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eOR (95% CI)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003ep-value\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003ctr\u003e \u003cth align=\"left\" colspan=\"3\" nameend=\"c3\" namest=\"c1\"\u003e \u003cp\u003eAgricultural land\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNo (Ref.)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eYes\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.38 (0.24\u0026ndash;0.61)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"3\" nameend=\"c3\" namest=\"c1\"\u003e \u003cp\u003e\u003cb\u003eExposed\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNo (Ref.)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eYes\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e6.65 (4.12\u0026ndash;10.73)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"3\" nameend=\"c3\" namest=\"c1\"\u003e \u003cp\u003e\u003cb\u003eLoss of agricultural land\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNo (Ref.)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eYes\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e6.23 (3.92\u0026ndash;9.88)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"3\" nameend=\"c3\" namest=\"c1\"\u003e \u003cp\u003e\u003cb\u003eLoss of cattle\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNo (Ref.)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eYes\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e5.17 (1.22\u0026ndash;21.92)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.026\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"3\" nameend=\"c3\" namest=\"c1\"\u003e \u003cp\u003e\u003cb\u003eMonthly earnings (taka)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u0026le;\u0026thinsp;6000\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e6001\u0026ndash;10000\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1.04 (0.42\u0026ndash;2.59)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.937\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e10001\u0026ndash;15000\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.63 (0.26\u0026ndash;1.50)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.293\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u0026ge;\u0026thinsp;15000\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.36 (0.15\u0026ndash;0.91)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.03\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"3\" nameend=\"c3\" namest=\"c1\"\u003e \u003cp\u003e\u003cb\u003eLoss of home\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNo (Ref.)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eYes\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e10.74 (5.72\u0026ndash;20.18)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"3\" nameend=\"c3\" namest=\"c1\"\u003e \u003cp\u003e\u003cb\u003eHope for land return\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNo (Ref.)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eYes\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e79504926.22 (0.00 - inf)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.998\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"3\" nameend=\"c3\" namest=\"c1\"\u003e \u003cp\u003e\u003cb\u003eNo of children\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e1\u0026ndash;2 (Ref.)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e3\u0026ndash;4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1.12 (0.73\u0026ndash;1.72)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.608\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u0026gt;\u0026thinsp;4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e3.32 (1.25\u0026ndash;8.82)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.016\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"3\" nameend=\"c3\" namest=\"c1\"\u003e \u003cp\u003e\u003cb\u003eSocial encouragement\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNo (Ref.)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eYes\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e3.66 (2.22\u0026ndash;6.02)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"3\" nameend=\"c3\" namest=\"c1\"\u003e \u003cp\u003e\u003cb\u003eSocial detachment\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNo (Ref.)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eYes\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e5.06 (2.74\u0026ndash;9.34)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"3\" nameend=\"c3\" namest=\"c1\"\u003e \u003cp\u003e\u003cb\u003eArea\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eKushtia (Ref.)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTangail\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e2.48 (1.33\u0026ndash;4.62)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.004\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRajbari\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1.90 (1.11\u0026ndash;3.25)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.019\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"3\" nameend=\"c3\" namest=\"c1\"\u003e \u003cp\u003e\u003cb\u003eAmount of agricultural land (decimal)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNo agricultural land (Ref.)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e1\u0026ndash;66\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.42 (0.25\u0026ndash;0.72)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e67\u0026ndash;99\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.29 (0.09\u0026ndash;0.98)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.046\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u0026gt;\u0026thinsp;100\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.30 (0.12\u0026ndash;0.74)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.009\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"3\" nameend=\"c3\" namest=\"c1\"\u003e \u003cp\u003e\u003cb\u003eEducational status\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eUneducated (Ref.)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eEducated\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.61 (0.40\u0026ndash;0.93)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.02\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"3\" nameend=\"c3\" namest=\"c1\"\u003e \u003cp\u003e\u003cb\u003eAge group (year)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e22\u0026ndash;37 (Ref.)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e38\u0026ndash;45\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.87 (0.54\u0026ndash;1.40)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.569\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e46\u0026ndash;80\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1.99 (1.15\u0026ndash;3.43)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.014\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"3\" nameend=\"c3\" namest=\"c1\"\u003e \u003cp\u003e\u003cb\u003eDuration of internal displacement (Years)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNot migrate\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u0026le;\u0026thinsp;3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e27.79 (9.94\u0026ndash;77.64)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u0026gt;\u0026thinsp;3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e2.66 (1.56\u0026ndash;4.52)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003cem\u003eRef: Reference\u003c/em\u003e.\u003c/p\u003e \u003cp\u003e \u003cb\u003eC. Hyperparameter Optimization of ML Algorithms and Comparative Analysis\u003c/b\u003e \u003c/p\u003e \u003cp\u003eTo complement the conventional statistical analysis, three supervised machine‑learning algorithms\u0026mdash;Logistic Regression (LR), Random Forest (RF), and Extreme Gradient Boosting (XGBoost)\u0026mdash;were trained to predict the risk of common mental health disorders (CMHD) in the study population. The final optimized hyperparameter configurations for each model are presented in Table\u0026nbsp;\u003cspan refid=\"Tab4\" class=\"InternalRef\"\u003e4\u003c/span\u003e.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab4\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 4\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eOptimized value of the Algorithms\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"2\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAlgorithms\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eOptimized values of Hyperparameters\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eLR\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e'C': 0.001, 'penalty': 'l2', 'solver': 'lbfgs\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRF\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e'max_depth': 10, 'max_features': 'log2', 'min_samples_split': 10, 'n_estimators': 500\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eXGBoost\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e'colsample_bytree': 0.8, 'learning_rate': 0.01, 'max_depth': 3, 'n_estimators': 200, 'subsample': 1.0\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003eModel performance was evaluated using Accuracy, Area Under the Receiver Operating Characteristic Curve (AUC), and F1‑score, with Precision and Recall additionally reported to characterize classification trade‑offs. The detailed comparative results of the optimized models are summarized in Table\u0026nbsp;\u003cspan refid=\"Tab5\" class=\"InternalRef\"\u003e5\u003c/span\u003e.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab5\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 5\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eComparative performance metrics of optimized machine learning algorithms.\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"6\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eModel\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eAccuracy\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eAUC Score\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eF1-Score\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003ePrecision\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003eRecall\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eLogistic Regression\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.762\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.746\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.865\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.762\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRandom Forest\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.776\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.757\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.863\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.807\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.926\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eXGBoost\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.771\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.763\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.864\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.788\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.957\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003eOverall, the three optimized models showed similar F1‑scores (0.863\u0026ndash;0.865), indicating a comparable balance between precision and recall in identifying CMHD in this dataset. The Random Forest model achieved the highest accuracy (0.776) and the highest precision (0.807), suggesting fewer false‑positive classifications relative to the other approaches. In contrast, XGBoost yielded the highest AUC (0.763), indicating the strongest overall discriminative ability across decision thresholds. Logistic Regression demonstrated perfect recall (1.000), meaning all CMHD cases were identified by the model, but this came with lower precision (0.762), consistent with a larger number of false positives. Taken together, the findings indicate that tree‑based ensemble models (RF and XGBoost) provided slightly stronger discrimination and/or precision, whereas Logistic Regression maximized sensitivity in this sample. Although XGBoost showed the highest AUC, Random Forest demonstrated the best overall balance across accuracy, precision, recall, and F1-score. Therefore, Random Forest was selected as the most appropriate algorithm for predicting CMHD, as shown in Figs.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e and \u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003e.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003cb\u003eD. Variable Significance and Feature Contribution using Random Forest\u003c/b\u003e \u003c/p\u003e \u003cp\u003eSince the Random Forest model showed the strongest overall predictive performance, we examined which predictors contributed most to its classification of CMHD risk. Feature contribution was quantified using the model\u0026rsquo;s built-in tree-based feature. To improve interpretability, importance values were aggregated at the variable level as presented in Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003e.\u003c/p\u003e \u003cp\u003eAs shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003e, house loss was by far the most influential predictor (importance\u0026thinsp;=\u0026thinsp;0.126), indicating that housing disruption is the dominant signal used by Random Forest when distinguishing individuals with CMHD from those without. The next most influential predictors were displacement (0.114) and land loss (0.101), highlighting the importance of erosion-related asset loss and mobility as major stress pathways. Exposure to erosion also contributed meaningfully (0.102), suggesting that direct environmental risk remains informative even after accounting for downstream losses. Beyond these primary shock-related factors, several household and psychosocial conditions contributed additional predictive information, including household income (0.099), number of child (0.058), and cope with that situation (cope_withh) (0.066). Overall, Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003e shows that Random Forest predictive power is driven mainly by displacement-related shocks\u0026mdash;especially house loss\u0026mdash;followed by land loss and displacement history, with exposure, number of child, coping, and income providing incremental contributions to CMHD risk classification.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e"},{"header":"4. Discussion","content":"\u003cp\u003eThis is a population based cross-sectional study, offer a representative estimate of the prevalence of CMHD and explores its association with individuals living in river erosion-prone areas. A comprehensive review of the literature found that no similar study has been conducted in the country. This study was designed to fill that gap by measuring the prevalence of CMHD and identifying the key risk factors contributing to these mental health issues in populations displaced or affected by riverbank erosion. Understanding these factors is crucial for developing targeted interventions and support programs to improve mental health outcomes in these vulnerable communities. However, our study found that people exposed to river erosion had significantly higher rates of CMHDs based on the DAS score compared to those not exposed (\u003cb\u003eFigure. 3\u003c/b\u003e). These results align with earlier studies [\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e, \u003cspan additionalcitationids=\"CR34\" citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e]. Thomas et al. [\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e] and Arobi et al. [\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e] found that people exposed to river erosion had a higher risk of developing DASS disorders compared to those not exposed.\u003c/p\u003e \u003cp\u003eThe reason is that exposed individuals lost various assets such as houses, property, and cattle, and some turned to drugs to cope with stress, anxiety, and frustration. Additionally, river erosion exposure may have worsened existing psychological distress, sleep problems, physical complaints, and psychosocial and behavioral issues [\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e]. We also found that the rates of abnormal Common Mental Health Disorders differed significantly among the three districts studied. Based on DAS scores, people in Rajbari and Tangail districts had a much higher risk of abnormal Common Mental Health Disorders compared to those in Kushtia district. One possible explanation is that during the dry season, people affected by erosion in Tangail and Rajbari districts had access to riverbeds along the Padma and Brahmaputra rivers due to their previous land boundaries. We also found that females were more likely to have DAS disorders than males, which coincided with earlier studies [\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e, \u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e]. According to DAS scores, Common Mental Health Disorders were significantly more common among illiterate people than those with education, consistent with findings from Altaf et al. and Weyerer et al. [\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e, \u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e]. Older age groups, especially those aged 38\u0026ndash;45 and above 45, had higher rates of abnormal Common Mental Health Disorders compared to younger people (\u0026le;\u0026thinsp;37 years), a pattern supported by previous research [\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e, \u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e]. In the study population, we found that the prevalence of abnormal CMHD increased as the number of children in a family grew. This may be because having more children adds to the economic burden caused by river erosion. Respondents with larger families had higher chances of developing CMHD compared to those with fewer children, as managing a big family under financial stress was especially difficult for affected individuals. This finding aligns with previous research [\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e]. Additionally, the odds of having abnormal CMHD were higher among landless participants living within 0.2 miles of the riverbank compared to those who owned land and lived farther than 0.2 miles from the river.\u003c/p\u003e \u003cp\u003eFurther, we use advanced machine learning (ML) techniques to greatly extend traditional statistical methods by capturing high-dimensional, non-linear interactions between socio-environmental stresses. With a peak accuracy of 0.776 and a precision of 0.807, the Random Forest model proved to be the best classifier, surpassing the baseline Logistic Regression (AUC\u0026thinsp;=\u0026thinsp;0.746). This discrepancy in performance suggests that complicated convergences, which are frequently oversimplified by traditional linear models, impact CMHD risk in populations affected by disasters.\u003c/p\u003e \u003cp\u003eThe feature importance analysis revealed \"Loss of House\" to be the main factor influencing CMHD classification, with the highest importance scores in the algorithmic framework, in line with the statistical evaluation. This finding indicates that the primary cause of psychological discomfort is the sudden loss of shelter. Algorithmic paths also demonstrated that secondary factors increase this risk; as the Lollipop Chart (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003e) illustrates, significant contributions from Land Ownership and Displacement together boost the model's predictive power. From the standpoint of public health, the Random Forest model's high Accuracy (0.776).\u003c/p\u003e \u003cp\u003eThese results show that ML-driven predictive modelling is an effective tool for policymakers, despite the drawbacks of cross-sectional data and the \"black-box\" character of ensemble models. After erosion incidents, it makes it easier to quickly identify high-risk homes, enabling more focused mental health interventions.\u003c/p\u003e \u003cp\u003e \u003cb\u003eStudy limitations\u003c/b\u003e \u003c/p\u003e \u003cp\u003eThis study had several limitations that should be considered when interpreting the findings. First, it was a cross-sectional study done in only three river erosion-prone districts (Kushtia, Rajbari, and Tangail) with a relatively small sample size, which means the findings may not apply to the whole country. Another limitation is that the study focused only on three major mental health conditions\u0026mdash;depression, anxiety, and stress\u0026mdash;as indicators of Common Mental Health Disorders (CMHD). Other important mental health issues, like neurodevelopmental disorders, sleep disorders, schizophrenia, bipolar disorder, obsessive-compulsive disorder, panic disorder, mixed anxiety and depression, and social phobia, were not covered in this study.\u003c/p\u003e"},{"header":"5. Conclusions","content":"\u003cp\u003eThis study demonstrates that people exposed to riverbank erosion had worse mental health and were more likely to develop common mental health disorders (CMHD) than those who weren\u0026rsquo;t exposed. Using advanced machine learning modeling, risk factors were identified, with the sudden \u0026ldquo;Loss of House\u0026rdquo;, \u0026ldquo;Loss of land\u0026rdquo;, and subsequent \u0026ldquo;Displacement\u0026rdquo; emerging as the most critical predictors of psychosocial distress. Additionally, our study highlights that certain demographic groups are more vulnerable to mental health disparities, such as women, the elderly, people without agricultural land, and those with bigger families (more than four children). Property loss and economic instability combine to create high-risk pathways for CMHD, as demonstrated by the Random Forest model's superior performance (Accuracy\u0026thinsp;=\u0026thinsp;0.776), which highlights the complex and non-linear nature of these socio-environmental stressors. Special attention should also be given to vulnerable groups, including women, the elderly, landless individuals, and those with larger families, to reduce mental health inequalities. Introducing low-interest loans in erosion-prone areas could further help recovery and improve people's ability to adapt.\u003c/p\u003e"},{"header":"Declarations","content":" \u003cp\u003e \u003cstrong\u003eEthical approval\u003c/strong\u003e \u003cp\u003eThe study\u0026rsquo;s procedures were conducted in accordance with the protocol approved by the Ethical Review Committee of Islamic University, Kushtia-7003, Bangladesh (Reference No. FS/IU/2025/17). All participants were informed prior to data collection about the study's objectives. The questionnaire was translated into the native language (Bangla) and read out to the respondents. When it was decided to conduct an interview, the educated respondents gave their written consent, and the uneducated participants signed the consent form with their fingerprints. Before taking part in this study, each participant gave their informed consent.\u003c/p\u003e \u003c/p\u003e \u003cp\u003e \u003cstrong\u003eConsent for publication\u003c/strong\u003e \u003cp\u003eNot applicable.\u003c/p\u003e \u003c/p\u003e \u003cp\u003e \u003cstrong\u003eCompeting interests\u003c/strong\u003e \u003cp\u003eThe authors declare no competing interests.\u003c/p\u003e \u003c/p\u003e\u003ch2\u003eFunding\u003c/h2\u003e \u003cp\u003eThe authors did not receive any specific funding for this study.\u003c/p\u003e\u003ch2\u003eAuthor Contribution\u003c/h2\u003e\u003cp\u003eConceptualization: Md. Tahidur Rahman and Md Lotifur Rohman; Data curation and statistical analysis: Md. Tahidur Rahman, Md Lotifur Rohman, Md. Maniruzzaman and Md. Merajul Islam; Drafting the manuscript: Md. Tahidur Rahman, Md Lotifur Rohman, Shyam Sundar Sarkar and Md. Fahim; Review and editing: Md. Tahidur Rahman, Md Lotifur Rohman, Md. Maniruzzaman and Md. Merajul Islam; Contribution to discussion, interpretation, and finalization: Md. Tahidur Rahman, Md Lotifur Rohman, Shyam Sundar Sarkar and Md. Fahim; Supervision: Md. Tahidur Rahman, Md. Maniruzzaman and Md. Merajul Islam; Validation. All authors have read and approved the final version of the manuscript.\u003c/p\u003e\u003ch2\u003eAcknowledgement\u003c/h2\u003e\u003cp\u003eThe authors would like to acknowledge to Dr. Md. Sazzed Hossain Zahid, Associate professor, Department of English, Islamic University Kushtia, Bangladesh, for his constructive feedback and assistance in improving the language of the manuscript.\u003c/p\u003e\u003ch2\u003eData Availability\u003c/h2\u003e\u003cp\u003eThe datasets used in the current study are available from the corresponding author upon reasonable request.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eLee EH, Moon SH, Cho MS, Park ES, Kim SY, Han JS, Cheio JH. The 21-Item and 12-Item Versions of the Depression Anxiety Stress Scales: Psychometric Evaluation in a Korean Population. Asian Nurs Res (Korean Soc Nurs Sci). 2019;13. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.anr.2018.11.006\u003c/span\u003e\u003cspan address=\"10.1016/j.anr.2018.11.006\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRapaport MH, Clary C, Fayyad R, Endicott J. Quality-of-life impairment in depressive and anxiety disorders. Am J Psychiatry. 2005;162. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1176/appi.ajp.162.6.1171\u003c/span\u003e\u003cspan address=\"10.1176/appi.ajp.162.6.1171\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLovibond SH, Lovibond PF. Manual for the Depression Anxiety Stress Scales. 1995.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWHO. Depressive disorder (depression). 2023.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBaxter AJ, Vos T, Scott KM, Norman RE, Flaxman AD, Blore J, Whiteford HA. The regional distribution of anxiety disorders: Implications for the Global Burden of Disease Study, 2010. Int J Methods Psychiatr Res. 2014;23. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1002/mpr.1444\u003c/span\u003e\u003cspan address=\"10.1002/mpr.1444\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNorris FH, Friedman MJ, Watson PJ, Byrne CM, Diaz E, Kaniasty K. 60,000 Disaster victims speak: Part I. An empirical review of the empirical literature, 1981\u0026ndash;2001. Psychiatry. 2002;65. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1521/psyc.65.3.207.20173\u003c/span\u003e\u003cspan address=\"10.1521/psyc.65.3.207.20173\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCostello A, Abbas M, Allen A, et al. Managing the health effects of climate change. Lancet and University College London Institute for Global Health Commission. Lancet. 2009;373. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/S0140-6736(09)60935-1\u003c/span\u003e\u003cspan address=\"10.1016/S0140-6736(09)60935-1\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWatts N, Amann M, Ayeb-Karlsson S, et al. The Lancet Countdown on health and climate change: from 25 years of inaction to a global transformation for public health. Lancet. 2018;391. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/S0140-6736(17)32464-9\u003c/span\u003e\u003cspan address=\"10.1016/S0140-6736(17)32464-9\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eOgbo FA, Mathsyaraja S, Koti RK, Perz J, Page A. The burden of depressive disorders in South Asia, 1990\u0026ndash;2016: findings from the global burden of disease study. BMC Psychiatry. 2018;18. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1186/s12888-018-1918-1\u003c/span\u003e\u003cspan address=\"10.1186/s12888-018-1918-1\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWHO: National Mental Health Survey of Bangladesh, 2018\u0026ndash;2019: provisional fact sheet. Dhaka. 2019. Published Online First: 2019.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMasum JH. Climatic Hazards in Bangladesh-A Literature Review Jahangir Hasan Masum. Published Online First. 2020. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.13140/RG.2.2.10239.94882\u003c/span\u003e\u003cspan address=\"10.13140/RG.2.2.10239.94882\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMunna TI. River erosion: A silent disaster in Bangladesh. 2019.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKamal M. Riverbank Erosion and Migration: A Study on Displaced People from Governance and Cultural Perspective. 2019, XIII:23\u0026ndash;34.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKhatun F, Bhattacharya D, Rahman M, Moazzem KG, Khan TI, Sabbih MA, Saadat S. Four Years of SDGs in Bangladesh: Measuring progress and charting the path forward. 2020.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eIPCC: Intergovernmental Panel on Climate Change working group II. Climate Change 2014: Impacts, Adaptation, and Vulnerability. Part B: Regional Aspects, Polar regions. New York: Cambridge Univ; 2014.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNagandla K, Nalliah S, Yin L, et al. Prevalence and associated risk factors of depression, anxiety and stress in pregnancy. Int J Reprod Contracept Obstet Gynecol Published Online First. 201610.18203/2320\u0026thinsp;\u0026ndash;\u0026thinsp;1770.ijrcog20162132.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMeltzer H, Bebbington P, Brugha T, Farrell M, Jenkins R. The relationship between personal debt and specific common mental disorders. Eur J Public Health. 2013;23:108\u0026ndash;13. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1093/eurpub/cks021\u003c/span\u003e\u003cspan address=\"10.1093/eurpub/cks021\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDarwish MA, Al Turki ZA, Sabra AA. Relation between Common Mental Disorders and Body Weight using Arabic DASS 21 scale (Depression, Anxiety and Stress Scale) among Adult Saudi Females Attending Primary Care. East Saudi Arabia. 2014.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSekoni O, Mall S, Christofides N. The relationship between protective factors and common mental disorders among female urban slum dwellers in Ibadan, Nigeria. PLoS ONE. 2022;17:1\u0026ndash;16. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1371/journal.pone.0263703\u003c/span\u003e\u003cspan address=\"10.1371/journal.pone.0263703\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNIH. Depression: Transforming the understanding and treatment of mental illnesses. 2024.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePark SG, Min KB, Chang SJ, Kim HC, Min JY. Job stress and depressive symptoms among Korean employees: The effects of culture on work. Int Arch Occup Environ Health. 2009;82. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1007/s00420-008-0347-8\u003c/span\u003e\u003cspan address=\"10.1007/s00420-008-0347-8\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eEske J. What is acute stress disorder? MedicalNewsToday. 2019.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePearson K. X. On the criterion that a given system of deviations from the probable in the case of a correlated system of variables is such that it can be reasonably supposed to have arisen from random sampling. London, Edinburgh, Dublin Philos Mag J Sci. 1900, 50:157\u0026ndash;75.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLavalley MP. Statistical Primer for Cardiovascular Research Logistic Regression. 2008, 2395\u0026ndash;9. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1161/CIRCULATIONAHA.106.682658\u003c/span\u003e\u003cspan address=\"10.1161/CIRCULATIONAHA.106.682658\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDesarrollo U, Desarrollo U, Desarrollo U. Integration of RNA Editing with Multiomics Data Improves Machine Learning Models for Predicting Drug Responses in Breast Cancer Patients. 2024.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChen T, Guestrin C. XGBoost: A scalable tree boosting system. Proc ACM SIGKDD Int Conf Knowl Discov Data Min. 2016, 13-17-Augu:785\u0026ndash;94. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1145/2939672.2939785\u003c/span\u003e\u003cspan address=\"10.1145/2939672.2939785\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLiashchynskyi P. Grid Search, Random Search, Genetic Algorithm: A Big Comparison for NAS. arXiv Prepr ArXiv191206059. 2019.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYang L, Shami A. On Hyperparameter Optimization of Machine Learning Algorithms: Theory and Practice. 1\u0026ndash;69.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSwets JA, Swets JA. Measuring the Accuracy of Diagnostic Systems Linked references are available on JSTOR for this article: Measuring the Accuracy of Diagnostic Systems. 1988, 240:1285\u0026ndash;93.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLobo JM, Jim\u0026eacute;nez-valverde A, Real R. AUC: a misleading measure of the performance of predictive distribution models. 2007, 1\u0026ndash;7. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1111/j.1466-8238.2007.00358.x\u003c/span\u003e\u003cspan address=\"10.1111/j.1466-8238.2007.00358.x\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSokolova M. Beyond Accuracy, F-score and ROC : a Family of Discriminant Measures for Performance Evaluation.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBuckland M, Gey F. The Relationship between Recall and Precision. Published Online First: 1994. 10.1002/(sici)1097\u0026ndash;4571(199401)45.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHossain A, Alam MJ, Haque MR. Effects of riverbank erosion on mental health of the affected people in Bangladesh. PLoS ONE. 2021;16. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1371/journal.pone.0254782\u003c/span\u003e\u003cspan address=\"10.1371/journal.pone.0254782\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMunro A, Kovats RS, Rubin GJ, et al. Effect of evacuation and displacement on the association between flooding and mental health outcomes: a cross-sectional analysis of UK survey data. Lancet Planet Heal. 2017;1. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/S2542-5196(17)30047-5\u003c/span\u003e\u003cspan address=\"10.1016/S2542-5196(17)30047-5\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJom Thomas J, Prakash B, Kulkarni P, Narayana Murthy MR. Prevalence and severity of depression among people residing in flood affected areas of Kerala. Int J Community Med Public Heal. 2019;6. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.18203/2394\u0026ndash;6040.ijcmph20190600\u003c/span\u003e\u003cspan address=\"10.18203/2394\u0026ndash;6040.ijcmph20190600\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSarzamArobi, Naher J, Soron TR. Impact of River Bank Erosion on Mental Health and Coping Capacity in Bangladesh. Glob Psychiatry. 2019;2. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.2478/gp-2019-0011\u003c/span\u003e\u003cspan address=\"10.2478/gp-2019-0011\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYeshaw Y, Mossie A. Depression, anxiety, stress, and their associated factors among Jimma university staff, Jimma, southwest Ethiopia, 2016: A cross-sectional study. Neuropsychiatr Dis Treat. 2017;13. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.2147/NDT.S150444\u003c/span\u003e\u003cspan address=\"10.2147/NDT.S150444\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWeyerer S, Eifflaender-Gorfer S, K\u0026ouml;hler L, et al. Prevalence and risk factors for depression in non-demented primary care attenders aged 75 years and older. J Affect Disord. 2008;111. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.jad.2008.02.008\u003c/span\u003e\u003cspan address=\"10.1016/j.jad.2008.02.008\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePinheiro KAT, Horta BL, Pinheiro RT, Horta LL, Terres NG, Da Silva RA. Common mental disorders in adolescents: A population based cross-sectional study. Rev Bras Psiquiatr. 2007;29. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1590/s1516-44462006005000040\u003c/span\u003e\u003cspan address=\"10.1590/s1516-44462006005000040\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"bmc-psychology","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"psyo","sideBox":"Learn more about [BMC Psychology](http://bmcpsychology.biomedcentral.com/)","snPcode":"","submissionUrl":"","title":"BMC Psychology","twitterHandle":"BMC_series","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"BMC Series","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"Depression, Anxiety, Stress, Riverbank erosion, Loss of house, Machine Learning, Bangladesh","lastPublishedDoi":"10.21203/rs.3.rs-9455372/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-9455372/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003ch2\u003eBackground\u003c/h2\u003e \u003cp\u003eRiverbank erosion is a recurrent environmental hazard in Bangladesh that leads to displacement, loss of livelihood, loss of land, and social instability. These disruptions place affected populations at heightened risk of common mental health disorders (CMHDs), including depression, anxiety, and stress. However, evidence on the relative importance of socioeconomic, environmental, and displacement-related determinants remains limited.\u003c/p\u003e\u003ch2\u003eObjective\u003c/h2\u003e \u003cp\u003eThis study aimed to identify key determinants of common mental health disorders among individuals living in riverbank erosion areas of Bangladesh using machine learning algorithms.\u003c/p\u003e\u003ch2\u003eMethods\u003c/h2\u003e \u003cp\u003eWe analyzed cross-sectional data collected from erosion-affected communities, where household heads were interviewed face-to-face between September 2021 and January 2022 to assess Common Mental Health Disorders (CMHDs) using the Depression, Anxiety, and Stress Scale (DASS-21) (Depression\u0026thinsp;\u0026ge;\u0026thinsp;10, Anxiety\u0026thinsp;\u0026ge;\u0026thinsp;8, or Stress\u0026thinsp;\u0026ge;\u0026thinsp;15). Descriptive statistics were conducted to determine the prevalence of CMHDs, while χ\u0026sup2;-test and logistic regression (LR) analysis were applied to identify statistically significant risk factors associated with CMHDs. Moreover, three optimized machine learning (ML) algorithms- Random Forest (RF), XGBoost, and LR were implemented for predicting CMHDs.\u003c/p\u003e\u003ch2\u003eResults\u003c/h2\u003e \u003cp\u003eA total of 711 households were included, and randomly allocated to training (70%; n\u0026thinsp;=\u0026thinsp;497) and testing (30%; n\u0026thinsp;=\u0026thinsp;214) datasets. Among 497 respondents, 398 (80.08%) were exposed to riverbank erosion. The prevalence of common mental health disorders was significantly higher among the exposed group than the non-exposed group (exposed: 84.17% versus non- exposed: 15.83%, P\u0026thinsp;\u0026lt;\u0026thinsp;0.001). The Random Forest classifier was found to be the best predictive model, with the accuracy (0.776) and precision (0.807), while loss of house, displacement, and loss of land were found to be the most influential predictors of CMHDs risk.\u003c/p\u003e\u003ch2\u003eConclusion\u003c/h2\u003e \u003cp\u003eThis study identified and predicted key risk factors for common mental health disorders among exposed individuals in riverbank erosion areas of Bangladesh using ML algorithms, which may assist policymakers in mitigating the burden of CMHDs, with particular attention to housing loss, displaced populations, and low-income individuals.\u003c/p\u003e","manuscriptTitle":"Identifying Associated Factors of Common Mental Health Disorders in Riverbank Erosion Areas of Bangladesh Using Machine Learning Algorithms","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-05-08 10:06:35","doi":"10.21203/rs.3.rs-9455372/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"reviewerAgreed","content":"151706520797468796252739301138856071246","date":"2026-05-11T02:49:52+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2026-04-30T03:57:22+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2026-04-28T23:47:44+00:00","index":"","fulltext":""},{"type":"editorInvited","content":"","date":"2026-04-27T04:43:55+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2026-04-25T09:29:12+00:00","index":"","fulltext":""},{"type":"submitted","content":"BMC Psychology","date":"2026-04-25T09:24:36+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"bmc-psychology","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"psyo","sideBox":"Learn more about [BMC Psychology](http://bmcpsychology.biomedcentral.com/)","snPcode":"","submissionUrl":"","title":"BMC Psychology","twitterHandle":"BMC_series","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"BMC Series","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"24832576-4b61-4484-a074-86c91178312c","owner":[],"postedDate":"May 8th, 2026","published":true,"recentEditorialEvents":[{"type":"reviewerAgreed","content":"151706520797468796252739301138856071246","date":"2026-05-11T02:49:52+00:00","index":39,"fulltext":""},{"type":"reviewersInvited","content":"18","date":"2026-04-30T03:57:22+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2026-04-28T23:47:44+00:00","index":"","fulltext":""}],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[],"tags":[],"updatedAt":"2026-05-08T10:06:36+00:00","versionOfRecord":[],"versionCreatedAt":"2026-05-08 10:06:35","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-9455372","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-9455372","identity":"rs-9455372","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}