Abstract
Background: Malaysia continues to face rising dengue incidence, particularly in high-density low-income housing where structural and behavioural vulnerabilities converge. This study evaluated the B.O.L.E.H. (B40 Communities Organising Local Solutions for Environmental Health) model—a multistage, community-governed intervention designed to strengthen resident capacity, improve knowledge, attitudes, and practices (KAP), and assess its association with trends in dengue incidence.Methods: A quasi-experimental pre–post design was implemented across four Program Perumahan Rakyat (PPR) communities in Lembah Pantai, Kuala Lumpur. Stage 1 involved advocate training and baseline KAP assessment at PPR Sri Pantai, followed by deployment of trained advocates to three neighbouring PPRs for Stage 2 implementation. The intervention included educational materials, lift-lobby posters, WhatsApp reminders, and champion-led engagement, which were evaluated via mosquito ovitraps and structured checklists to measure changes in environmental risk and community participation. KAP surveys were conducted at baseline, midline, and post-intervention. Monthly dengue incidence was obtained from national surveillance and analysed using negative binomial Generalised Estimating Equations with a log-population offset to estimate group, time, and interaction effects.Findings: Intervention communities demonstrated significant improvements in knowledge and practice (p<0·001). Monthly dengue incidence declined 5% per month in intervention localities (IRR 0·95, 0·92–0·98; p=0·004). Control communities experienced more than three-fold higher incidence overall (IRR 3·11, 2·56–3·78; p<0·0001) and a significantly faster increase in incidence over time (time×group IRR 1·05, 1·04–1·06; p<0·0001). In a separate predictors model, at the PPR level, higher mean knowledge scores were associated with lower dengue incidence (IRR 0·42 per one-point increase in mean knowledge; 95% CI 0·36–0·50).Interpretation: Embedding resident-led micro-governance through floor-level leadership, routine monitoring, and recognition systems was associated with sustained preventive action and a more favourable dengue-incidence trajectory. Given the non-randomised design, the small number of PPRs, baseline socioeconomic differences, and the ecological nature of the predictors analysis, these findings cannot establish causality. However, they suggest that the B.O.L.E.H. model could usefully complement national dengue-control strategies as a community-governed layer in high-risk urban settings.