Changing household dietary behaviours through community-based networks: A pragmatic cluster randomized controlled trial in rural Kerala, India

Trial design: With the rise in prevalence of non-communicable diseases in India and Kerala in particular, efforts to develop lifestyle interventions have increased. However, contextualised interventions are limited. We developed and implemented contextualised behavioural intervention strategies focusing on household dietary behaviours in selected rural areas in Kerala and conducted a community-based pragmatic cluster randomized controlled trial to assess its effectiveness to increase the intake of fruits and vegetables at individual level, and the procurement of fruits and vegetables at the household level and reduce the consumption of salt, sugar and oil at the household level. Methods: Six out of 22 administrative units in the northern part of Thiruvananthapuram district of Kerala state were selected as geographic boundaries and randomized to either intervention or control arms. Stratified sampling was carried out and 30 clusters comprising 6–11 households were selected in each arm. A cluster was defined as a neighbourhood group functioning in rural areas under a state-sponsored community-based network (Kudumbasree). We screened 1237 households and recruited 479 (intervention: 240; control: 239) households and individuals (male or female aged 25–45 years) across the 60 clusters. 471 households and individuals completed the intervention and end-line survey and one was excluded due to pregnancy. Interventions were delivered for a period of one-year at household level at 0, 6, and 12 months, including counselling sessions, telephonic reminders, home visits and general awareness sessions through the respective neighbourhood groups in the intervention arm. Households in the control arm received general dietary information leaflets. Data from 478 households (239 in each arm) were included in the intention-to-treat analysis, with the household as the unit of analysis. Results: There was significant, modest increase in fruit intake from baseline in the intervention arm (12.5%); but no significant impact of the intervention on vegetable intake over the control arm. There was a significant increase in vegetable procurement in the intervention arm compared to the control arm with the actual effect size showing an overall increase by19%; 34% of all households in the intervention arm had increased their procurement by at least 20%, compared to 17% in the control arm. Monthly household consumption of salt, sugar and oil was greatly reduced in the intervention arm compared to the control arm with the actual effect sizes showing an overall reduction by 45%, 40% and 48% respectively. Conclusions: The intervention enabled significant reduction in salt, sugar and oil consumption and improvement in fruit and vegetable procurement at the household level in the intervention arm. However, there was a disconnect between the demonstrated increase in FV procurement and the lack of increase in FV intake. We need to explore fruit and vegetable intake behaviour further to identify strategies or components that would have made a difference. We can take forward the lessons learned from this study to improve our understanding of human dietary behaviour and how that can be changed to improve health within this context.