Factors associated with obesity among female workers in small and cottage industries in Bangladesh: a cross-sectional study Authors: Nusrat Jahan Mim, Tasnim Tamanna, Rafath Farah Rashid, Shabnam Ara Jahan, Shazia Afrin, Nasrin Akter, Bilkis Banu Published: 2026 ABSTRACT BackgroundFemales in low- and middle-income countries are vulnerable to obesity due to biological factors such as fat distribution and reproductive physiology and social constraints including limited physical activity and caregiving roles. In Bangladesh, occupational research has largely focused on ready-made garment workers, while females in small and cottage industries such as shrimp processing, handloom, pottery, and nakshi kantha remain underrepresented despite their substantial workforce contribution. This study aimed to estimate the prevalence of obesity among female workers in selected small and cottage industries and to examine occupational, lifestyle, and dietary factors associated with obesity.MethodsA cross-sectional study was conducted between May 2023 and February 2024 among 350 female workers aged 18–60 years across four small and cottage industry sectors (handloom, shrimp farming, pottery, and nakshi kantha) in selected districts of Bangladesh. Study sites were purposively selected to represent major production clusters, and participants were recruited consecutively within sites, limiting generalizability. Data were collected via face-to-face interviews using a semi-structured questionnaire. BMI was measured by trained nutritionists, with obesity defined as ≥ 25 kg/m² (WHO South-East Asia). Associations were examined using chi-square tests and multivariable binary logistic regression analyses with a priori covariate selection, reporting crude odds ratios (COR) and adjusted odds ratios (AOR) with 95% confidence intervals (CIs).ResultsObesity prevalence was 36.0%. In the multivariable analysis, factors significantly associated with obesity included age 31–45 years (AOR 1.979, 95% CI: 1.026–3.819), sedentary work (AOR 2.954, 95% CI: 1.363–6.400), sitting ≥ 8 h/day (AOR 4.233, 95% CI: 2.147–8.346), gastrointestinal complaints (AOR 3.180, 95% CI: 1.237–8.178), physical inactivity (AOR 2.478, 95% CI: 1.131–5.429), screen time ≥ 2 h/day (AOR 3.750, 95% CI: 1.983–7.092), and daily soft drink and energy drink consumption (AOR 2.884, 95% CI: 1.196–6.955). Model fit was acceptable (Hosmer–Lemeshow p = 0.070; Omnibus test p < 0.001; Nagelkerke R² = 0.397).ConclusionsObesity was common and associated with modifiable occupational and lifestyle factors rather than industry type. These findings suggest potential value in workplace-oriented interventions targeting sedentary behavior, physical inactivity, and unhealthy beverage consumption in similar informal work settings. Background Workplace health promotion is recognized as a key public health priority, as the workplace substantially shapes adult health through the time spent in occupational settings [1]. Female workers may be particularly vulnerable due to the dual burden of productive and reproductive responsibilities [2]. Obesity is a major global public health concern that disproportionately affects females, who experience increased metabolic, cardiovascular, and reproductive complications [3]. Biological factors (e.g., fat distribution and reproductive physiology) and social determinants (e.g., limited physical activity, gendered work roles, and caregiving responsibilities) further increase females’ risk of obesity [4]. Using Asian-specific cut-offs recommended for South-East Asian populations, a body mass index (BMI) ≥ 25 kg/m² is considered indicative of obesity. In Bangladesh, recent estimates indicate that 30% of adults were affected in 2022, with a higher prevalence among females than males (36% vs. 20%) [4]. Comparable patterns have also been reported in occupational settings in neighboring countries, including 25% of female handloom workers in India and 48.6% of female workers in Indonesian palm plantations presenting with BMI ≥ 25 kg/m² [5, 6]. Obesity among workers reflects interacting occupational, behavioral, dietary, and social factors that jointly influence energy balance [7]. Occupational exposures such as task type, physical demands, and working posture, together with sedentary behavior, insufficient physical activity, and dietary patterns, act through interconnected pathways. These risks are often shaped by broader socioeconomic conditions, particularly in informal labor settings where female workers balance income-generating activities with caregiving responsibilities under constrained resources [8, 9]. In low- and middle-income countries, such transitions have contributed to rising obesity prevalence and associated non-communicable disease risks, along with reduced functional capacity, productivity loss, and increased absenteeism and disability [10–12]. Within this context, small and cottage industries in Bangladesh represent an important yet understudied occupational setting. These industries are heterogeneous, encompassing food processing, textiles, pottery, and handicrafts, and involve a notable proportion of female workers [13]. In shrimp processing, females comprise approximately 80% of the casual labor force and perform repetitive tasks such as cleaning, beheading, freezing, and packaging [14]. In handloom work, females constitute about half of the workforce and engage in prolonged sitting and repetitive manual activities [15]. In pottery, females represent approximately 40% of workers and perform shaping, drying, and finishing tasks involving static postures and moderate physical effort [16, 17]. In handicrafts such as nakshi kantha, work is predominantly sedentary and often home- or workshop-based [18]. Despite its public health importance, research on occupational obesity in Bangladesh has largely focused on ready-made garment workers [10, 19]. Evidence from small and cottage industries remains limited, particularly regarding studies applying a multifactorial analytical approach integrating occupational exposures, dietary behaviors, and lifestyle factors. In addition, research specifically addressing female workers’ occupational conditions and dual work–care responsibilities is scarce. Therefore, this study aims to estimate the prevalence of obesity among female workers in selected small and cottage industries and to identify associated occupational, lifestyle, and dietary determinants. The findings will inform the development of gender-sensitive, sector-specific workplace health strategies. Methods Study design and setting This cross-sectional study, conducted between May 2023 and February 2024, aimed to estimate the prevalence of obesity and its associated factors among female workers in selected small and cottage industries in Bangladesh. The study included four sectors: handloom (textile and leather), shrimp farming (food, drinks, and tobacco processing), pottery (chemical and petroleum), and nakshi kantha (handicraft), selected due to their high concentration of female employment and economic importance in their respective regions [13–18]. The study sites were purposively selected to represent major production clusters for each sector, as summarized in Table 1. Tangail district in Dhaka Division, a historic center for handloom production, contributes over 11% of national output, with women comprising approximately 50% of the workforce [15]. Bagerhat and Satkhira districts in Khulna Division were selected for shrimp farming, which together account for approximately 75% of the country’s shrimp farming area, with women constituting around 80% of casual labor in shrimp processing plants [14]. Kakran Pottery Village in Dhaka District consists of approximately 8–10 workshops employing both men and women, of whom approximately 40% are women [16, 17]. Chapainawabganj district in Rajshahi Division was selected for the nakshi kantha sector, where approximately 300 production units employ about 60,000 workers, 95% of whom are women [20]. Table 1Summary of selected sectors, subsectors, study sites, and female workforceSectorSelected sub sectorDivisionDistrictsFemale (%)Textile and leatherHandloomDhakaTangail50Food, drinks, and tobacco processingShrimp farmingKhulnaBagerhat, Satkhira80Chemical and petroleumPotteryDhakaDhaka40HandicraftNakshi kanthaRajshahiChapainawabganj95 Table 1Summary of selected sectors, subsectors, study sites, and female workforceSectorSelected sub sectorDivisionDistrictsFemale (%)Textile and leatherHandloomDhakaTangail50Food, drinks, and tobacco processingShrimp farmingKhulnaBagerhat, Satkhira80Chemical and petroleumPotteryDhakaDhaka40HandicraftNakshi kanthaRajshahiChapainawabganj95 Summary of selected sectors, subsectors, study sites, and female workforce Participants and sampling Purposive sampling was employed to include major production clusters where female workers are geographically concentrated. Within each selected site, participants were recruited using a consecutive sampling approach. No formal sampling frame was available for these informal-sector workers. Eligible participants were female workers aged 18–60 years, reflecting the typical working age in these sectors and aligning with prior studies [21]. Participants were required to be currently employed in one of the selected industries and willing to provide informed consent. Exclusion criteria included incomplete or inconsistent responses; pregnancy or lactation; or a self-reported history of severe psychiatric illness. Sample size The required sample size was determined using Cochran’s formula: n = \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\frac{{Z}^{2}pq}{{d}^{2}}$$\end{document}, where Z = 1.96, p = 0.353 (prevalence of BMI ≥ 25 kg/m² among female workers [19]), q = 1 − p, and d = 0.05. The minimum required sample size was 350. To account for potential non-response, the sample size was inflated, and 400 eligible participants were approached. Of these, 350 completed the survey, yielding a response rate of 87.5%. No design effect was applied, as the study used a non-probability sampling design with purposively selected sites and consecutive recruitment within those sites, rather than a formal cluster sampling framework. Data collection A preliminary pilot study involving five participants from each site was conducted to assess questionnaire clarity and feasibility, and minor modifications were made accordingly. Data were collected using a semi-structured questionnaire administered face-to-face in Bengali by trained interviewers. Interviews were conducted between May 2023 and February 2024 and lasted approximately 15–20 min per participant. At each site, data collectors were present throughout the work shift on scheduled data collection days. All workers encountered during this period were screened for eligibility, and those meeting inclusion criteria were recruited consecutively. Recruitment continued throughout the workday, including both active work periods and breaks, until the required sample size was achieved. The questionnaire was developed based on relevant literature and established guidelines [11, 21–32]. To minimize measurement error, trained interviewers administered the questionnaire using standardized protocols, and responses were checked immediately after completion for completeness and internal consistency. Questionnaires with incomplete or inconsistent responses were excluded from the final analysis. As a result, the dataset used for analysis contained no missing values. Local administrative authorities facilitated access to workplaces and coordination with participants. Ethical considerations The study was approved by the Ethical Review Committee of Northern University Bangladesh, Dhaka, Bangladesh (NUB/DPH/EC/2023/28), and conducted in accordance with the Declaration of Helsinki. Participation was anonymous and voluntary, and written informed consent was obtained from all participants. Participants were free to withdraw at any time. Variables Sociodemographic characteristics Twelve covariates were selected a priori for inclusion in the multivariable binary logistic regression model based on theoretical relevance, prior evidence, and to ensure an appropriate events-per-variable ratio. Sociodemographic variables included age and monthly family income. Age was categorized as 18–30, 31–45, and 46–60 years [21]. Monthly family income was categorized as ≤ 6,000 BDT ( ≈ ≤ USD 55.6), 6,001–8,000 BDT (≈ USD 55.6–74.1), 8,001–10,000 BDT (≈ USD 74.1–92.6), and 10,001–12,000 BDT (≈ USD 92.6–111.1), based on the 2023 average exchange rate (1 USD ≈ 108 BDT) [22, 33]. Occupational characteristics Occupational variables included current working sector, work pattern, average daily sitting time, and self-reported occupational health problems. Current working sector was categorized as handloom, shrimp farming, pottery, or nakshi kantha. Work pattern was categorized as sedentary craft or manual/active based on primary work tasks [23]. Average daily sitting time was categorized as < 8 or ≥ 8 h [24]. Self-reported occupational health problems during the past year were collected using an open-ended question and classified as none, musculoskeletal, respiratory, gastrointestinal, or others [25]. Although related, work pattern, average sitting time, and physical activity represent distinct and non-equivalent dimensions of exposure, capturing occupational posture and task characteristics (work pattern), sedentary duration during work (sitting time), and overall movement behavior across occupational and non-occupational contexts (physical activity). Lifestyle and dietary characteristics Lifestyle variables included physical activity, sleep duration, and daily screen time. Physical activity was assessed using a single-item question aligned with World Health Organization recommendations and categorized as sufficiently active, insufficiently active, or inactive [34]. Sleep duration was dichotomized as < 7 or ≥ 7 h per day [26]. Daily screen time was categorized as < 2 or ≥ 2 h per day [27, 28]. Dietary variables included breakfast consumption, starchy roots and tubers intake, and soft drink and energy drink consumption, all of which were self-reported. Breakfast frequency was assessed with the question, “How often do you eat breakfast?” with response options: almost every day, sometimes, rarely, or never. Participants reporting breakfast sometimes, rarely, or never were classified as having irregular breakfast consumption, consistent with prior studies linking irregular breakfast patterns to obesity risk [29–31]. This categorization facilitated comparability with existing literature. Dietary intake was assessed using a reduced, frequency-based approach adapted from food frequency questionnaire (FFQ) methodology previously validated in Bangladeshi populations [32]. To maintain model parsimony and focus on obesity-related dietary exposures, two domains were retained: a vegetable-based component and beverage intake [35–37]. The vegetable component was operationalized using selected items from the “starchy roots, tubers and their products” group of the Bangladesh food composition table, including commonly consumed items such as potato, sweet potato, and colocasia/taro and related tubers (e.g., elephant foot yam, giant taro, yam), typically consumed boiled with or without salt [38]. Item selection was informed by local dietary relevance and consultation with nutritionists. Soft drink and energy drink consumption was assessed using a single item capturing intake of carbonated and commercially sweetened beverages [32]. Consumption frequency was initially recorded using six response categories [39] and subsequently collapsed into four categories (never; 1–3 times per month; 1–6 times per week; and daily) to ensure adequate cell sizes and stable model estimation. Anthropometric measurements Body weight and height were measured by trained nutritionists following standard protocols. Body weight was measured to the nearest 0.1 kg using a calibrated digital scale (Model RGZ-160, China), with participants barefoot, wearing light clothing, and after voiding [11]. Height was measured to the nearest 0.1 cm using a stadiometer, with participants standing barefoot in the Frankfurt plane [11]. Body mass index (BMI) was calculated as weight (kg) divided by height squared (m²) and classified according to World Health Organization South-East Asian guidelines, with obesity defined as BMI ≥ 25 kg/m² [11]. This lower cut-off was used because South Asian populations, including Bangladesh, experience higher cardiometabolic risk at lower BMI levels, and regional evidence shows that global cut-offs may underestimate obesity-related health risks in this setting [11]. All measurements were conducted under standardized conditions by trained nutritionists to ensure accuracy and reproducibility. Statistical analysis Data were entered and analyzed using IBM SPSS Statistics version 27. No missing data were present in the final dataset; therefore, no imputation procedures were required. Categorical variables were summarized using frequencies and percentages, while continuous variables (e.g., age, monthly family income) were described using means and standard deviations. The prevalence of obesity was calculated as the proportion of participants with BMI ≥ 25 kg/m², based on World Health Organization South-East Asia Regional Office guidelines. Differences in sociodemographic, occupational, lifestyle, and dietary characteristics across industries were assessed using chi-square tests. Bivariate associations between obesity and independent variables were initially examined using chi-square tests. The conceptualization of obesity in this study was guided by a socioecological framework, recognizing the interaction of distal sociodemographic factors, intermediate occupational exposures, and proximal behavioral and dietary determinants. Confounders were selected a priori based on theoretical considerations and prior literature, and not selected based on bivariate statistical significance. All selected covariates were included in the multivariable analysis irrespective of their bivariate significance. Binary logistic regression was performed to identify factors associated with obesity. Crude odds ratios (COR) with 95% confidence intervals (CI) were estimated from univariate models, and adjusted odds ratios (AOR) with 95% CIs were obtained from the multivariable model. All twelve independent variables were included in the multivariable binary logistic regression model to ensure adequate control of confounding and to maintain an appropriate events-per-variable ratio. No interaction terms were included, and effect modification was not assessed. Variables were conceptually grouped as distal sociodemographic factors (age, monthly income), intermediate occupational factors (work pattern, sitting duration, working sector, and occupational health problems), and proximal lifestyle and dietary behaviors (physical activity, sleep duration, screen time, breakfast consumption, and dietary intake). This grouping was based on conceptual distinctions between occupational and lifestyle-related variables rather than statistical associations, to preserve separation of exposure domains in the model. All variables were entered simultaneously in the multivariable binary logistic regression model to estimate independent associations with obesity, acknowledging potential interrelationships across domains due to overlapping behavioral and occupational exposures. Model adequacy was evaluated using the Hosmer–Lemeshow goodness-of-fit test, the Omnibus test of model coefficients, and Nagelkerke R². Multicollinearity was assessed using variance inflation factors (VIF) and tolerance values, with thresholds of VIF < 5 and tolerance ≥ 0.20 indicating acceptable levels. Statistical significance was set at p < 0.05. The study followed the STROBE Statement for reporting cross-sectional studies [40]. Results Descriptive profile of the study participants A total of 350 female workers from four cottage industry sectors were included in the analysis. The mean age was 34.25 ± 10.50 years, with 43.4% aged 18–30 years, 40.3% aged 31–45 years, and 16.3% aged 46–60 years. The mean monthly family income was BDT 9730 ± 2778.72, and 51.7% of participants reported an income of BDT 10,001–12,000. Regarding occupational characteristics, 66.6% were engaged in sedentary craft-based work and 59.4% reported sitting for ≥ 8 h per day at work. Respiratory conditions (30.9%) were the most frequently reported occupational health complaint. Lifestyle characteristics showed that 44.9% of participants were physically inactive, 29.1% reported sleeping < 7 h per day, and 66.0% reported daily screen time ≥ 2 h. Irregular breakfast consumption was reported by 80.6%. For dietary intake, 39.4% reported consuming starchy roots and tubers 1–3 times per month, while 34.6% reported daily soft drink and energy drink consumption (Table 2). Distribution of participants’ characteristics by working sector Significant differences across working sectors were observed in sociodemographic (age and income), occupational (work pattern, sitting duration, and health complaints), lifestyle (physical activity, screen time, and breakfast consumption), and dietary characteristics (starchy food intake and soft drink consumption) (all p < 0.001), whereas sleep duration did not differ significantly across sectors (p = 0.085). Sociodemographic characteristics varied across working sectors. Age distribution differed significantly by industry (χ² = 42.396, p < 0.001), with a higher proportion of younger workers in handloom and older workers in pottery. Monthly family income also differed across sectors (χ² = 36.361, p < 0.001), with relatively higher representation of higher-income groups in the handloom sector and lower-income categories more represented in pottery. Occupational characteristics showed statistically significant variation across industries. Work pattern differed by sector (χ² = 79.723, p < 0.001), with sedentary craft work more common in handloom and manual/active work more frequent in shrimp farming. Sitting duration at work also varied (χ² = 34.096, p < 0.001), with ≥ 8 h/day sitting more frequently observed in the handloom sector. Occupational health complaints differed across sectors (χ² = 56.702, p < 0.001). Respiratory complaints were more frequently reported in handloom, musculoskeletal complaints in shrimp farming, and gastrointestinal complaints in nakshi kantha. Lifestyle-related characteristics also varied significantly by sector. Physical activity frequency differed across industries (χ² = 45.053, p < 0.001), with a higher proportion of inactive participants in handloom. Screen time ≥ 2 h/day was more common in handloom (χ² = 33.524, p < 0.001). Breakfast consumption patterns also differed across sectors (χ² = 20.071, p < 0.001), with irregular intake more frequent in handloom. Dietary intake varied significantly by working sector. Consumption of starchy roots and tubers differed across industries (χ² = 50.557, p < 0.001), with lower-frequency intake more common in handloom. Soft drink and energy drink consumption also differed across sectors (χ² = 64.077, p < 0.001), with higher daily consumption reported in handloom (Table 2). Table 2Distribution of participants’ characteristics by working sector (n = 350)VariablesCurrent working sectorTotal, n (%)HI, n (%)SFI, n (%)PI, n (%)NKI, n (%)\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\boldsymbol{X}}^{2}$$\end{document}/pSociodemographic characteristicsAge (years) (34.25 ± 10.50) Young (18–30)152 (43.4)76 (50.0)11 (7.2)23 (15.1)42 (27.6)42.396/<0.001 Middle (31–45)141 (40.3)48 (34.0)26 (18.4)43 (30.5)24 (17.0) Older (46–60)57 (16.3)14 (24.6)16 (28.1)22 (38.6)5 (8.8)Monthly family incomes (BDT) (9730 ± 2778.72) ≤ 600060 (17.1)9 (15.0)14 (23.3)21 (35.0)16 (26.7)36.361/<0.001 6001–800055 (15.7)14 (25.5)12 (21.8)13 (23.6)16 (29.1) 8001–10,00054 (15.4)22 (40.7)10 (18.5)10 (18.5)12 (22.2) 10,001–12,000181 (51.7)93 (51.4)17 (9.4)44 (24.3)27 (14.9)Occupational characteristicsWork pattern Manual/active117 (33.4)21 (17.9)47 (40.2)33 (28.2)16 (13.7)79.723/<0.001 Sedentary craft233 (66.6)117 (50.2)6 (2.6)55 (23.6)55 (23.6)Sitting hours at work (per day) < 8 h/day142 (40.6)31 (21.8)40 (28.2)57 (40.1)14 (9.9)34.096/<0.001 ≥ 8 h/day208 (59.4)107 (51.4)13 (6.3)31 (14.9)57 (27.4)Primary self-reported occupational health problem (past year) None115 (32.9)57 (49.6)10 (8.7)38 (33.0)10 (8.7)56.702/<0.001 Musculoskeletal65 (18.6)19 (29.2)20 (30.8)17 (26.2)9 (13.8) Respiratory108 (30.9)43 (39.8)10 (9.3)22 (20.4)33 (30.6) Gastrointestinal39 (11.1)10 (25.6)5 (12.8)11 (28.2)13 (33.3) Others23 (6.6)9 (39.1)8 (34.8)0 (0.0)6 (26.1)Lifestyle and dietary characteristicsWeekly physical activity frequency Sufficiently active94 (26.9)36 (38.3)14 (14.9)34 (36.2)10 (10.6)45.053/<0.001 Insufficiently active99 (28.3)30 (30.3)22 (22.2)35 (35.4)12 (12.1) Inactive157 (44.9)72 (45.9)17 (10.8)19 (12.1)49 (31.2)Daily sleep duration (hours) < 7 h102 (29.1)32 (31.4)17 (16.7)34 (33.3)19 (18.6)6.626/0.085 ≥ 7 h248 (70.9)106 (42.7)36 (14.5)54 (21.8)52 (21.0)Daily overall screen time (hours) < 2 h119 (34.0)42 (35.3)35 (29.4)17 (14.3)25 (21.0)33.524/<0.001 ≥ 2 h231 (66.0)96 (41.6)18 (7.8)71 (30.7)46 (19.9)Daily breakfast consumption frequency Regular68 (19.4)19 (27.9)20 (29.4)22 (32.4)7 (10.3)20.071/<0.001 Irregular282 (80.6)119 (42.2)33 (11.7)66 (23.4)64 (22.7)Starchy roots and tubers consumption frequency Never85 (24.3)41 (48.2)13 (15.3)20 (23.5)11 (12.9)50.557/<0.001 1–3 times/month138 (39.4)66 (47.8)7 (5.1)35 (25.4)30 (21.7) 1–6 times/week65 (18.6)10 (15.4)25 (38.5)16 (24.6)14 (21.5) Everyday62 (17.7)21 (33.9)8 (12.9)17 (27.4)16 (25.8)Soft drink and energy drink consumption frequency Never63 (18.0)35 (55.6)15 (23.8)6 (9.5)7 (11.1)64.077/<0.001 1–3 times/month118 (33.7)29 (24.6)11 (9.3)39 (33.1)39 (33.1) 1–6 times/week48 (13.7)20 (41.7)17 (35.4)5 (10.4)6 (12.5) Everyday121 (34.6)54 (44.6)10 (8.3)38 (31.4)19 (15.7)aHI Handloom industry, SFI Shrimp farming industry, PI Pottery industry, NKI Nakshi kantha industry, χ² Chi-square test, n number of participants, p p-valuep < 0.05 indicates statistical significance; bolded p-values represent statistical significance Table 2Distribution of participants’ characteristics by working sector (n = 350)VariablesCurrent working sectorTotal, n (%)HI, n (%)SFI, n (%)PI, n (%)NKI, n (%)\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\boldsymbol{X}}^{2}$$\end{document}/pSociodemographic characteristicsAge (years) (34.25 ± 10.50) Young (18–30)152 (43.4)76 (50.0)11 (7.2)23 (15.1)42 (27.6)42.396/<0.001 Middle (31–45)141 (40.3)48 (34.0)26 (18.4)43 (30.5)24 (17.0) Older (46–60)57 (16.3)14 (24.6)16 (28.1)22 (38.6)5 (8.8)Monthly family incomes (BDT) (9730 ± 2778.72) ≤ 600060 (17.1)9 (15.0)14 (23.3)21 (35.0)16 (26.7)36.361/<0.001 6001–800055 (15.7)14 (25.5)12 (21.8)13 (23.6)16 (29.1) 8001–10,00054 (15.4)22 (40.7)10 (18.5)10 (18.5)12 (22.2) 10,001–12,000181 (51.7)93 (51.4)17 (9.4)44 (24.3)27 (14.9)Occupational characteristicsWork pattern Manual/active117 (33.4)21 (17.9)47 (40.2)33 (28.2)16 (13.7)79.723/<0.001 Sedentary craft233 (66.6)117 (50.2)6 (2.6)55 (23.6)55 (23.6)Sitting hours at work (per day) < 8 h/day142 (40.6)31 (21.8)40 (28.2)57 (40.1)14 (9.9)34.096/<0.001 ≥ 8 h/day208 (59.4)107 (51.4)13 (6.3)31 (14.9)57 (27.4)Primary self-reported occupational health problem (past year) None115 (32.9)57 (49.6)10 (8.7)38 (33.0)10 (8.7)56.702/<0.001 Musculoskeletal65 (18.6)19 (29.2)20 (30.8)17 (26.2)9 (13.8) Respiratory108 (30.9)43 (39.8)10 (9.3)22 (20.4)33 (30.6) Gastrointestinal39 (11.1)10 (25.6)5 (12.8)11 (28.2)13 (33.3) Others23 (6.6)9 (39.1)8 (34.8)0 (0.0)6 (26.1)Lifestyle and dietary characteristicsWeekly physical activity frequency Sufficiently active94 (26.9)36 (38.3)14 (14.9)34 (36.2)10 (10.6)45.053/<0.001 Insufficiently active99 (28.3)30 (30.3)22 (22.2)35 (35.4)12 (12.1) Inactive157 (44.9)72 (45.9)17 (10.8)19 (12.1)49 (31.2)Daily sleep duration (hours) < 7 h102 (29.1)32 (31.4)17 (16.7)34 (33.3)19 (18.6)6.626/0.085 ≥ 7 h248 (70.9)106 (42.7)36 (14.5)54 (21.8)52 (21.0)Daily overall screen time (hours) < 2 h119 (34.0)42 (35.3)35 (29.4)17 (14.3)25 (21.0)33.524/<0.001 ≥ 2 h231 (66.0)96 (41.6)18 (7.8)71 (30.7)46 (19.9)Daily breakfast consumption frequency Regular68 (19.4)19 (27.9)20 (29.4)22 (32.4)7 (10.3)20.071/<0.001 Irregular282 (80.6)119 (42.2)33 (11.7)66 (23.4)64 (22.7)Starchy roots and tubers consumption frequency Never85 (24.3)41 (48.2)13 (15.3)20 (23.5)11 (12.9)50.557/<0.001 1–3 times/month138 (39.4)66 (47.8)7 (5.1)35 (25.4)30 (21.7) 1–6 times/week65 (18.6)10 (15.4)25 (38.5)16 (24.6)14 (21.5) Everyday62 (17.7)21 (33.9)8 (12.9)17 (27.4)16 (25.8)Soft drink and energy drink consumption frequency Never63 (18.0)35 (55.6)15 (23.8)6 (9.5)7 (11.1)64.077/<0.001 1–3 times/month118 (33.7)29 (24.6)11 (9.3)39 (33.1)39 (33.1) 1–6 times/week48 (13.7)20 (41.7)17 (35.4)5 (10.4)6 (12.5) Everyday121 (34.6)54 (44.6)10 (8.3)38 (31.4)19 (15.7)aHI Handloom industry, SFI Shrimp farming industry, PI Pottery industry, NKI Nakshi kantha industry, χ² Chi-square test, n number of participants, p p-valuep < 0.05 indicates statistical significance; bolded p-values represent statistical significance Distribution of participants’ characteristics by working sector (n = 350) aHI Handloom industry, SFI Shrimp farming industry, PI Pottery industry, NKI Nakshi kantha industry, χ² Chi-square test, n number of participants, p p-value p < 0.05 indicates statistical significance; bolded p-values represent statistical significance Prevalence of obesity in the total sample and by industry Among the 350 female workers, the overall prevalence of obesity (BMI ≥ 25 kg/m², South-East Asian cut-off) was 36.0%, while 64.0% were non-obese. By industry, obesity prevalence differed across sectors, being highest among handloom workers (41.3%), followed by nakshi kantha (36.6%) and pottery (31.8%), and lowest among shrimp farming workers (28.3%). The distribution of obesity across industries was examined using a chi-square test, which indicated no statistically significant association between industry type and obesity status (χ² = 3.728, p = 0.292). Accordingly, these findings are presented for descriptive comparison only (Fig. 1). Fig. 1Distribution of obesity (BMI ≥ 25 kg/m², South-East Asian cut-off) among female workers in small and cottage industries in Bangladesh, overall and by industry (handloom, shrimp farming, pottery, and nakshi kantha) Distribution of obesity (BMI ≥ 25 kg/m², South-East Asian cut-off) among female workers in small and cottage industries in Bangladesh, overall and by industry (handloom, shrimp farming, pottery, and nakshi kantha) Figure 1. Distribution of obesity (BMI ≥ 25 kg/m², South-East Asian cut-off) among female workers in small and cottage industries in Bangladesh, overall and by industry (handloom, shrimp farming, pottery, and nakshi kantha). Bivariate associations between studied variables and obesity In bivariate analyses, several occupational, lifestyle, and dietary factors were significantly associated with obesity, whereas no significant associations were observed for age, monthly family income, or current working sector (p > 0.05). All studied variables were included in the multivariable binary logistic regression model to identify independent associations with obesity (Table 3). Table 3Association between studied variables and obesity status (n = 350)VariablesObesity status (obese BMI = ≥ 25 kg/m²)Obese, n (%)Non-obese, n (%)\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\boldsymbol{X}}^{2}$$\end{document}/pAge (years) Young (18–30)47 (30.9)105 (69.1)4.506/0.105 Middle (31–45)60 (42.6)81 (57.4) Older (46–60)19 (33.3)38 (66.7)Monthly family incomes (in BDT) ≤ 600021 (35.0)39 (65.0)0.456/0.928 6001–800022 (40.0)33 (60.0) 8001–10,00019 (35.2)35 (64.8) 10,001–12,00064 (35.4)117 (64.6)Current working sector Handloom57 (41.3)81 (58.7)3.728/0.292 Shrimp farming15 (28.3)38 (71.7) Pottery28 (31.8)60 (68.2) Nakshi kantha26 (36.6)45 (63.4)Work pattern Manual/active25 (21.4)92 (78.6)16.332/<0.001 Sedentary craft101 (43.3)132 (56.7)Sitting hours at work (per day) < 8 h/day29 (20.4)113 (79.6)25.165/<0.001 ≥ 8 h/day97 (46.6)111 (53.4)Primary self-reported occupational health problem (past year) None40 (34.8)75 (65.2)15.662/0.004 Musculoskeletal16 (24.6)49 (75.4) Respiratory36 (33.3)72 (66.7) Gastrointestinal24 (61.5)15 (38.5) Others10 (43.5)13 (56.5)Weekly physical activity frequency Sufficiently active34 (36.2)60 (63.8)8.001/0.018 Insufficiently active25 (25.3)74 (74.7) Inactive67 (42.7)90 (57.3)Daily sleep duration (hours) < 7 h38 (37.3)64 (62.7)0.098/0.754 ≥ 7 h88 (35.5)160 (64.5)Daily overall screen time (hours) < 2 h23 (19.3)96 (80.7)21.753/<0.001 ≥ 2 h103 (44.6)128 (55.4)Daily breakfast consumption frequency Regular15 (22.1)53 (77.9)7.119/0.008 Irregular111 (39.4)171 (60.6)Starchy roots and tubers consumption frequency Never40 (47.1)45 (52.9)10.592/0.014 1–3 times/month38 (27.5)100 (72.5) 1–6 times/week28 (43.1)37 (56.9) Everyday20 (32.3)42 (67.7)Soft drink and energy drink consumption frequency Never16 (25.4)47 (74.6)18.605/<0.001 1–3 times/month30 (25.4)88 (74.6) 1–6 times/week21 (43.8)27 (56.3) Everyday59 (48.8)62 (51.2)aχ² = chi-square test, n number of participants, BMI Body mass index, p p-valuep < 0.05 indicates statistical significance; bolded p-values represent statistical significance Table 3Association between studied variables and obesity status (n = 350)VariablesObesity status (obese BMI = ≥ 25 kg/m²)Obese, n (%)Non-obese, n (%)\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\boldsymbol{X}}^{2}$$\end{document}/pAge (years) Young (18–30)47 (30.9)105 (69.1)4.506/0.105 Middle (31–45)60 (42.6)81 (57.4) Older (46–60)19 (33.3)38 (66.7)Monthly family incomes (in BDT) ≤ 600021 (35.0)39 (65.0)0.456/0.928 6001–800022 (40.0)33 (60.0) 8001–10,00019 (35.2)35 (64.8) 10,001–12,00064 (35.4)117 (64.6)Current working sector Handloom57 (41.3)81 (58.7)3.728/0.292 Shrimp farming15 (28.3)38 (71.7) Pottery28 (31.8)60 (68.2) Nakshi kantha26 (36.6)45 (63.4)Work pattern Manual/active25 (21.4)92 (78.6)16.332/<0.001 Sedentary craft101 (43.3)132 (56.7)Sitting hours at work (per day) < 8 h/day29 (20.4)113 (79.6)25.165/<0.001 ≥ 8 h/day97 (46.6)111 (53.4)Primary self-reported occupational health problem (past year) None40 (34.8)75 (65.2)15.662/0.004 Musculoskeletal16 (24.6)49 (75.4) Respiratory36 (33.3)72 (66.7) Gastrointestinal24 (61.5)15 (38.5) Others10 (43.5)13 (56.5)Weekly physical activity frequency Sufficiently active34 (36.2)60 (63.8)8.001/0.018 Insufficiently active25 (25.3)74 (74.7) Inactive67 (42.7)90 (57.3)Daily sleep duration (hours) < 7 h38 (37.3)64 (62.7)0.098/0.754 ≥ 7 h88 (35.5)160 (64.5)Daily overall screen time (hours) < 2 h23 (19.3)96 (80.7)21.753/<0.001 ≥ 2 h103 (44.6)128 (55.4)Daily breakfast consumption frequency Regular15 (22.1)53 (77.9)7.119/0.008 Irregular111 (39.4)171 (60.6)Starchy roots and tubers consumption frequency Never40 (47.1)45 (52.9)10.592/0.014 1–3 times/month38 (27.5)100 (72.5) 1–6 times/week28 (43.1)37 (56.9) Everyday20 (32.3)42 (67.7)Soft drink and energy drink consumption frequency Never16 (25.4)47 (74.6)18.605/<0.001 1–3 times/month30 (25.4)88 (74.6) 1–6 times/week21 (43.8)27 (56.3) Everyday59 (48.8)62 (51.2)aχ² = chi-square test, n number of participants, BMI Body mass index, p p-valuep < 0.05 indicates statistical significance; bolded p-values represent statistical significance Association between studied variables and obesity status (n = 350) aχ² = chi-square test, n number of participants, BMI Body mass index, p p-value p < 0.05 indicates statistical significance; bolded p-values represent statistical significance Multivariable binary logistic regression analysis of factors associated with obesity A total of 350 female workers from small and cottage industries were included in the analysis. In the adjusted model, participants aged 31–45 years had higher odds of obesity compared to those aged 18–30 years (AOR = 1.979, 95% CI: 1.026–3.819, p = 0.042), while no significant association was observed for those aged 46–60 years. Occupational factors were strongly associated with obesity. Workers engaged in sedentary craft work had higher odds of obesity compared to those in manual or active work (AOR = 2.954, 95% CI: 1.363–6.400, p = 0.006). Similarly, sitting duration of ≥ 8 h per day was associated with increased odds of obesity (AOR = 4.233, 95% CI: 2.147–8.346, p < 0.001). Participants reporting physical inactivity also had higher odds of obesity compared to those who were sufficiently active (AOR = 2.478, 95% CI: 1.131–5.429, p = 0.007). Self-reported gastrointestinal problems were associated with higher odds of obesity (AOR = 3.180, 95% CI: 1.237–8.178, p = 0.016), while musculoskeletal and respiratory conditions were not statistically significant after adjustment. Lifestyle behaviors showed significant associations. Higher screen time (≥ 2 h/day) was associated with increased odds of obesity (AOR = 3.750, 95% CI: 1.983–7.092, p < 0.001). Participants who reported everyday consumption of soft drinks and energy drinks had higher odds of obesity compared to non-consumers (AOR = 2.884, 95% CI: 1.196–6.955, p = 0.018). No significant associations were observed for breakfast frequency, sleep duration, or starchy roots and tubers intake after adjustment. The final model showed acceptable fit (Hosmer–Lemeshow test: p = 0.070; Omnibus test: p < 0.001) with moderate explanatory power (Nagelkerke R² = 0.397). This level of explanatory power is considered moderate within behavioral epidemiological models and is likely influenced by the co-occurrence and interrelated nature of occupational and lifestyle-related exposures in the study population rather than model overfitting. No evidence of multicollinearity was observed (VIF range: 1.049–1.182; tolerance range: 0.846–0.954). The model included 12 predictors with 126 outcome events, yielding an events-per-variable ratio of 10.5 (Table 4). Table 4Crude and adjusted odds ratios for factors associated with obesity among female workers (n = 350)VariablesObesity status (BMI)COR (95% CI) p AOR (95% CI) p Age (years) Young (18–30)Ref.Ref. Middle (31–45)1.655 (1.025–2.673) 0.039 1.979 (1.026–3.819) 0.042 Older (46–60)1.117 (0.584-2.138)0.7380.875 (0.369-2.074)0.762Monthly family incomes (in BDT) ≤ 6000Ref.Ref. 6001–80001.238 (0.581-2.638)0.5801.520 (0.578-3.999)0.396 8001–10,0001.008 (0.467-2.178)0.9831.070 (0.419-2.735)0.887 10,001–12,0001.016 (0.551-1.873)0.9600.766 (0.346 − 1.700)0.513Current working sector HandloomRef.Ref. Shrimp farming0.561 (0.282-1.115)0.0992.473 (0.687-8.906)0.166 Pottery0.663 (0.378-1.163)0.1521.210 (0.513-2.854)0.664 Nakshi kantha0.821 (0.455-1.481)0.5120.708 (0.313-1.598)0.405Work pattern Manual/activeRef.Ref. Sedentary craft2.816 (1.687-4.700) < 0.001 2.954 (1.363-6.400) 0.006 Sitting hours at work (per day) < 8 h/dayRef.Ref. ≥ 8 h/day3.405 (2.085–5.561) < 0.001 4.233 (2.147–8.346) < 0.001 Primary self-reported occupational health problem (past year) NoneRef.Ref. Musculoskeletal0.612 (0.309-1.211)0.1590.495 (0.188-1.302)0.154 Respiratory0.937 (0.539-1.632)0.8200.745 (0.320-1.735)0.495 Gastrointestinal3.000 (1.416–6.355) 0.004 3.180 (1.237–8.178) 0.016 Others1.442 (0.581-3.581)0.4301.069 (0.308-3.712)0.916Weekly physical activity frequency Sufficiently activeRef.Ref. Insufficiently active0.596 (0.321-1.107)0.1010.902 (0.397-2.049)0.857 Inactive1.314 (0.776-2.224)0.3102.478 (1.131–5.429) 0.007 Daily sleep duration (hours) < 7 hRef.Ref. ≥ 7 h0.926 (0.574-1.494)0.7540.900 (0.488-1.662)0.737Daily overall screen time (hours) < 2 hRef.Ref. ≥ 2 h3.359 (1.989–5.671) < 0.001 3.750 (1.983–7.092) < 0.001 Daily breakfast consumption frequency RegularRef.Ref. Irregular2.294 (1.233–4.268) 0.009 1.300 (0.585 − 2.890)0.520Starchy roots and tubers consumption frequency Never1.867 (0.944-3.691)0.0731.611 (0.664-3.911)0.292 1–3 times/month0.798 (0.416-1.529)0.4970.658 (0.284-1.527)0.330 1–6 times/week1.589 (0.770-3.279)0.2102.058 (0.725-5.842)0.175 EverydayRef.Ref.Soft drink and energy drink consumption frequency NeverRef.Ref. 1–3 times/month1.001 (0.496-2.022)0.9970.877 (0.341-2.257)0.785 1–6 times/week2.285 (1.022–5.107) 0.044 1.691 (0.559-5.118)0.353 Everydy2.795 (1.430–5.463) 0.003 2.884 (1.196–6.955) 0.018 aCOR Crude odds ratio, AOR Adjusted odds ratio, CI Confidence interval, BMI Body mass index, Ref Reference category, p p-valuebBMI reference category = < 25 kg/m² (non-obese, coded 0, reference), obese = ≥ 25 kg/m² (coded 1, outcome); adjusted model includes all studied variables; adjusted model fitness assessed by Hosmer–Lemeshow test: χ² = 14.495, p = 0.070; Omnibus test: χ² = 119.576, p < 0.001; Nagelkerke R² = 0.397; p < 0.05 indicates statistical significance; bolded p-values represent statistical significance Table 4Crude and adjusted odds ratios for factors associated with obesity among female workers (n = 350)VariablesObesity status (BMI)COR (95% CI) p AOR (95% CI) p Age (years) Young (18–30)Ref.Ref. Middle (31–45)1.655 (1.025–2.673) 0.039 1.979 (1.026–3.819) 0.042 Older (46–60)1.117 (0.584-2.138)0.7380.875 (0.369-2.074)0.762Monthly family incomes (in BDT) ≤ 6000Ref.Ref. 6001–80001.238 (0.581-2.638)0.5801.520 (0.578-3.999)0.396 8001–10,0001.008 (0.467-2.178)0.9831.070 (0.419-2.735)0.887 10,001–12,0001.016 (0.551-1.873)0.9600.766 (0.346 − 1.700)0.513Current working sector HandloomRef.Ref. Shrimp farming0.561 (0.282-1.115)0.0992.473 (0.687-8.906)0.166 Pottery0.663 (0.378-1.163)0.1521.210 (0.513-2.854)0.664 Nakshi kantha0.821 (0.455-1.481)0.5120.708 (0.313-1.598)0.405Work pattern Manual/activeRef.Ref. Sedentary craft2.816 (1.687-4.700) < 0.001 2.954 (1.363-6.400) 0.006 Sitting hours at work (per day) < 8 h/dayRef.Ref. ≥ 8 h/day3.405 (2.085–5.561) < 0.001 4.233 (2.147–8.346) < 0.001 Primary self-reported occupational health problem (past year) NoneRef.Ref. Musculoskeletal0.612 (0.309-1.211)0.1590.495 (0.188-1.302)0.154 Respiratory0.937 (0.539-1.632)0.8200.745 (0.320-1.735)0.495 Gastrointestinal3.000 (1.416–6.355) 0.004 3.180 (1.237–8.178) 0.016 Others1.442 (0.581-3.581)0.4301.069 (0.308-3.712)0.916Weekly physical activity frequency Sufficiently activeRef.Ref. Insufficiently active0.596 (0.321-1.107)0.1010.902 (0.397-2.049)0.857 Inactive1.314 (0.776-2.224)0.3102.478 (1.131–5.429) 0.007 Daily sleep duration (hours) < 7 hRef.Ref. ≥ 7 h0.926 (0.574-1.494)0.7540.900 (0.488-1.662)0.737Daily overall screen time (hours) < 2 hRef.Ref. ≥ 2 h3.359 (1.989–5.671) < 0.001 3.750 (1.983–7.092) < 0.001 Daily breakfast consumption frequency RegularRef.Ref. Irregular2.294 (1.233–4.268) 0.009 1.300 (0.585 − 2.890)0.520Starchy roots and tubers consumption frequency Never1.867 (0.944-3.691)0.0731.611 (0.664-3.911)0.292 1–3 times/month0.798 (0.416-1.529)0.4970.658 (0.284-1.527)0.330 1–6 times/week1.589 (0.770-3.279)0.2102.058 (0.725-5.842)0.175 EverydayRef.Ref.Soft drink and energy drink consumption frequency NeverRef.Ref. 1–3 times/month1.001 (0.496-2.022)0.9970.877 (0.341-2.257)0.785 1–6 times/week2.285 (1.022–5.107) 0.044 1.691 (0.559-5.118)0.353 Everydy2.795 (1.430–5.463) 0.003 2.884 (1.196–6.955) 0.018 aCOR Crude odds ratio, AOR Adjusted odds ratio, CI Confidence interval, BMI Body mass index, Ref Reference category, p p-valuebBMI reference category = < 25 kg/m² (non-obese, coded 0, reference), obese = ≥ 25 kg/m² (coded 1, outcome); adjusted model includes all studied variables; adjusted model fitness assessed by Hosmer–Lemeshow test: χ² = 14.495, p = 0.070; Omnibus test: χ² = 119.576, p < 0.001; Nagelkerke R² = 0.397; p < 0.05 indicates statistical significance; bolded p-values represent statistical significance Crude and adjusted odds ratios for factors associated with obesity among female workers (n = 350) aCOR Crude odds ratio, AOR Adjusted odds ratio, CI Confidence interval, BMI Body mass index, Ref Reference category, p p-value bBMI reference category = < 25 kg/m² (non-obese, coded 0, reference), obese = ≥ 25 kg/m² (coded 1, outcome); adjusted model includes all studied variables; adjusted model fitness assessed by Hosmer–Lemeshow test: χ² = 14.495, p = 0.070; Omnibus test: χ² = 119.576, p < 0.001; Nagelkerke R² = 0.397; p < 0.05 indicates statistical significance; bolded p-values represent statistical significance Discussion In this study of 350 female workers across four small and cottage industry sectors in Bangladesh, the overall prevalence of obesity (BMI ≥ 25 kg/m²) was 36.0%, indicating a substantial burden of excess weight in this occupationally underrepresented population. This estimate is broadly comparable to findings among female garment workers in Bangladesh (35.3%) [19], but lower than the 56% prevalence reported among ready-made garment workers using Asian BMI cut-offs [10]. Regional studies also show considerable variability, including 25% prevalence of BMI 25.0–29.9 among female handloom workers in India [5] and 48.6% obesity among female plantation workers in Indonesia [6]. Such variation likely reflects differences in occupational structure, sex composition, local food environments, physical workload, measurement cut-offs, and sampling strategies. In addition, obesity prevalence can vary according to the population studied, the extent of sedentary work, and whether studies use WHO standard or Asian-specific BMI thresholds. Despite this variability, the present findings reinforce that women in informal and low-wage labor sectors carry a substantial burden of excess adiposity and related cardiometabolic risk [12]. To our knowledge, this is the first study to examine obesity across multiple cottage industry sectors simultaneously in Bangladesh while also considering occupational, lifestyle, and dietary correlates. Previous research in Bangladesh has largely focused on rural populations [3, 41], clinical female samples [14], male cottage industry workers [22], or single-industry settings for non-obesity outcomes [14]. The present findings therefore add sector-comparative evidence from a population that has received little attention in national obesity research. Age showed a clear association with obesity, with women aged 31–45 years having nearly twice the odds of obesity compared with those aged 18–30 years. This pattern is consistent with national and regional evidence showing higher obesity prevalence among women in mid-adult age groups [3, 4]. A plausible explanation is the accumulation of life-course exposures, including reproductive history, hormonal changes, and declining activity patterns, alongside prolonged exposure to sedentary occupational conditions. Similar age gradients have been reported in other Asian settings [42], supporting the relevance of age-sensitive prevention strategies in similar settings. Occupational sedentary behavior was strongly associated with obesity. Females engaged in sedentary work and those reporting ≥ 8 h/day of sitting had higher odds of obesity, consistent with evidence linking prolonged sitting with greater adiposity and cardiometabolic risk [23, 24]. This finding suggests that workplace organization may be an important determinant of energy balance in these sectors. Limited movement opportunities during work may contribute to reduced energy expenditure, particularly in industries where tasks are repetitive and posture is maintained for long periods. Screen-based sedentary behavior and physical inactivity were also positively associated with obesity. Women reporting ≥ 2 h/day of screen time had higher odds of obesity, in line with studies linking screen exposure to increased adiposity [27, 28]. This relationship may reflect co-occurrence of behaviors such as reduced physical activity, increased exposure to food marketing, concurrent snacking, and irregular sleep patterns [28]. Unlike occupational sitting, screen time represents a modifiable leisure behavior, suggesting potential relevance for behavioral health promotion strategies. Physical inactivity was associated with approximately two and a half times higher odds of obesity, consistent with extensive evidence showing inverse relationships between physical activity and adiposity [43]. This pattern may be influenced by interactions between occupational constraints and dietary intake. Evidence suggests that physical activity patterns are more effective in influencing weight status when considered alongside dietary behaviors [43]. Together, these findings suggest that both occupational and non-occupational sedentary patterns may contribute to obesity risk in this population. Gastrointestinal complaints were associated with approximately threefold higher odds of obesity. Similar associations have been reported previously [25, 44], although the direction of this relationship cannot be determined from cross-sectional data. Gastrointestinal symptoms may coexist with dietary patterns, occupational stressors, or other health conditions that cluster with obesity. A bidirectional relationship is also plausible, as obesity may influence gastrointestinal function through metabolic and inflammatory mechanisms, while gastrointestinal discomfort may affect eating behaviors. Longitudinal studies are needed to clarify temporal ordering. Dietary behavior showed selective associations. Daily consumption of soft drinks and energy drinks was associated with nearly threefold higher odds of obesity, consistent with evidence linking sugar-sweetened beverages with overweight and obesity [35]. These beverages are characterized by high sugar content and low satiety, which may influence overall energy intake patterns [35]. Nonetheless, they may support consideration of nutrition-oriented public health strategies, including taxation and product reformulation approaches [45]. These beverages contribute energy without promoting satiety and may therefore increase total caloric intake. By contrast, starchy roots and tubers intake was not independently associated with obesity after adjustment, suggesting that the crude association may have been confounded by other occupational or behavioral factors. Working sector was not independently associated with obesity after adjustment, indicating that the crude differences observed across industries were more likely explained by underlying behavioral and occupational factors than by sector type itself. This suggests that obesity risk may be shaped more by working conditions, movement patterns, and beverage consumption than by industry classification alone. In practical terms, interventions may be more effective if directed toward modifiable working conditions and daily behaviors rather than sector identity. To support worker health, workplaces may also consider strengthening access to basic occupational health services, safety practices, regular health monitoring, and health education, alongside attention to working conditions such as working hours and occupational well-being [19]. Policy responses may further benefit from life-course and gender-sensitive perspectives, particularly given the higher likelihood of obesity observed among women in mid-adulthood in this and other studies [4]. This study has several strengths. It included a relatively large sample of 350 female workers from four major cottage industry sectors, improving coverage of key informal occupational groups. Anthropometric data were collected by trained personnel using standardized procedures, which strengthened measurement reliability. The use of multivariable binary logistic regression with a broad set of a priori selected covariates allowed adjustment for multiple potential confounders. Model diagnostics indicated acceptable fit (Hosmer–Lemeshow p = 0.070), and there was no evidence of problematic multicollinearity. Reporting also followed STROBE guidelines. Several limitations should be considered. The cross-sectional design prevents causal inference and limits interpretation of temporality. Purposive site selection, consecutive recruitment, and the absence of a formal sampling frame may restrict generalizability beyond similar production clusters. Several behavioral and dietary measures were self-reported, so recall and reporting bias cannot be ruled out. Physical activity was measured with a single item, which may not have captured overall activity patterns fully. Dietary assessment was simplified and frequency based, so it did not assess total energy intake, portion sizes, or full macronutrient composition; collapsing categories may also have reduced variability and limited assessment of dose–response patterns. Residual confounding from unmeasured factors remains possible, and some shared variance among occupational, behavioral, and dietary predictors may have influenced the estimates. Future research should examine whether the observed associations vary across subgroups by testing for effect modification. Prospective or longitudinal studies would also help clarify temporal relationships between workplace conditions, lifestyle factors, and obesity-related outcomes. In addition, intervention studies are needed to assess the feasibility and effectiveness of integrated workplace programs addressing nutrition, physical activity, and occupational health in similar informal-sector settings. Conclusion The study found a high prevalence of obesity (36.0%) among female workers in Bangladesh’s small and cottage industries. In the multivariable binary logistic regression analysis, obesity was independently associated with age, occupational sedentary behavior, prolonged sitting time, physical inactivity, screen-based sedentary behavior, gastrointestinal complaints, and daily consumption of soft drinks and energy drinks. No independent associations were observed for industry type or most dietary variables after adjustment. Overall, these findings indicate that obesity in this population is associated with a combination of occupational and lifestyle-related factors within informal work settings. Given the cross-sectional design, these associations should not be interpreted as causal, but they may help guide further analytical and longitudinal research in similar populations.