Prevalence and cow-level risk factors for mild and severe subclinical mastitis in intensive and extensive dairy systems in Bangladesh.

Subclinical mastitis (SCM) substantially limits dairy productivity in Bangladesh, yet evidence from high-density herds remains limited. This cross-sectional study estimated the prevalence of mild and severe SCM and identified cow-level risk factors in 3,173 lactating cows from 76 herds in Sirajganj District, Bangladesh (January 2023-December 2024). Composite milk samples were tested using an EKOMILK SCAN® analyzer to measure somatic cell count (SCC), and cows were classified as healthy, mild SCM, or severe SCM using SCC distribution-based thresholds. Cow-level data were collected using structured questionnaires. A Bayesian hierarchical multinomial regression model, with herd as a random effect, was used to estimate relative risk ratios (RRR). Overall, 1,517 cows (47.8%) had mild SCM and 578 cows (18.2%) had severe SCM, with comparable prevalence between extensive and intensive systems (66.4% vs 65.9%; chi-square p = 0.836). Lactation stage showed severity-specific effects: late lactation credibly increased mild SCM risk (RRR 1.758; 95% CrI 1.306-2.365) but not severe SCM risk, whereas early lactation credibly increased severe SCM risk (RRR 1.595). History of clinical mastitis was the strongest predictor of severity: relative to cows with two prior episodes, severe SCM risk was higher with one episode (RRR 5.754; 95% CrI 2.709-12.533), three or more episodes (RRR 5.162), or no prior history (RRR 2.131), similarly for mild SCM. Lower parity and higher milk yield credibly reduced SCM risk; Holstein Friesian and Jersey crosses had credibly lower mild SCM risk than Sahiwal crosses, and pregnant cows had credibly higher mild SCM risk. Targeted surveillance of late-lactation and mastitis-history cows may improve SCM control in high-density herds.