Pathways linking to behavioural and biological risk factors on diabetes and overweight/obese among adults in urban slums from Western India: a structural equation modelling analysis

The prevalence of non-communicable diseases (NCDs) is on the rise in India, exhibiting significant regional disparities. This study aimed to evaluate sociodemographic, behavioural, biological and anthropometric risk factors associated with diabetes and overweight/obese among adults residing in urban slums of western India. This cross-sectional community-based study in India involved 456 adults from urban slums, selected through three-stage sampling. Data collection was done by using the WHO STEPS approach, with participants providing informed consent and completing a pretested questionnaire covering sociodemographic, behavioral, and biological risk factors. They were interviewed, and anthropometric measurements were done using standardized procedures and fasting blood samples were analysed for biochemical parameters using standard laboratory methods. Results were expressed as proportions and analysis was done by using multivariate logistic regression, one-way ANOVA, & post hoc Bonferroni test. Structural equation modelling (SEM) was applied to evaluate direct and indirect associations involving Diabetes and BMI outcome- overweight/obese. The prevalence of Diabetes, hypertension and overweight/obese was 13.2%, 14.9%, and 28.1%, respectively. Behavioural risk factors included alcohol consumption (13.8%), smoking (23.2%), inadequate fruit consumption (51.3%), inadequate vegetable consumption (98.9%), and insufficient physical activity (78.7%). Biological risk factors included hypercholesterolaemia (19.5%), hypertriglyceridaemia (21.3%), and hyperhomocysteinaemia (69.5%). In multivariable logistic regression, age > 40 years (AOR 3.57, 95% CI: 1.60–7.96) was independently associated with Diabetes, while participants in BJ Prasad socioeconomic Class III had lower odds of Diabetes (AOR 0.24, 95% CI: 0.08–0.73). Smoking (AOR 3.22, 95% CI: 1.62–6.38), alcohol consumption (AOR 2.24, 95% CI: 1.03–4.89), and hyperhomocysteinaemia (AOR 2.35, 95% CI: 1.15–4.82) were independently associated with hypertension, whereas fruit consumption was associated with lower odds (AOR 0.37, 95% CI: 0.20–0.69). Insufficient physical activity was independently associated with overweight/obese (AOR 1.87, 95% CI: 1.11–3.15). The SEM demonstrated excellent model fit and identified significant associations of age (β = 0.168, p < 0.001), smoking (β = 0.143, p < 0.001), alcohol consumption (β = 0.108, p < 0.01), hypercholesterolemia (β = 0.097, p < 0.05), and socioeconomic status (β = 0.092, p < 0.05) with Diabetes, while physical inactivity showed a significant direct association with overweight/obese (β = 0.105, p < 0.05) and LDL cholesterol ( β = +0.121, p = 0.009) were significantly and positively associated with BMI overweight/obese. The findings highlight the importance of strengthening NCD surveillance and preventive health programmes in urban slum populations.