We conducted a longitudinal dietary intervention study to assess the impact of a store-based intervention on mediators and moderators and consequent dietary behaviour in Indigenous communities in remote Australia. We assessed dietary intake of fruit, vegetable, water and sweetened soft drink, mediators and moderators among 148, eighty-five and seventy-three adult participants (92 % women) at baseline (T1), end of intervention (T2) and at 24 weeks post intervention (T3), respectively. Mediators included perceived affordability and self-efficacy. Moderators were barriers to eat more fruit and vegetables and food security. Mixed-effects models were used to determine changes in mediators and moderators with time and associations between these and each dietary outcome. Perceived vegetable affordability increased from T1 (19 %; 95 % CI 11, 27) to T2 (38 %; 95 % CI 25, 51) (P=0·004) and returned to baseline levels at T3. High self-efficacy to eat more fruit and vegetables and to drink less soft drink decreased from T1 to T3. A reduction in soft drink intake of 27 % (95 % CI -44, -4; P=0·02) was reported at T3 compared with T1; no changes with time were observed for all other outcome measures. Regardless of time, vegetable intake was positively associated with self-efficacy to cook and try new vegetables, no barriers and food security. The dietary intervention went someway to improving perceived affordability of vegetables but was probably not strong enough to overcome other mediators and moderators constraining behaviour change. Meaningful dietary improvement in this context will be difficult to achieve without addressing underlying constraints to behaviour change.
Keywords: Dietary intakes; Diets; Indigenous communities; Longitudinal sub-studies; Mediators; Moderators.