Background Small Island Developing States (SIDS) in the Caribbean and Pacific have high burdens of nutrition related disease, including type 2 diabetes, cardiovascular diseases, and other non-communicable diseases (NCDs), micronutrient deficiencies and in several countries substantial, persistent, childhood stunting and wasting. These burdens are associated with diets high in nutrient poor, energy dense, foods. This highlights the need for evidence-based interventions. We aimed to systematically review interventions intended to improve nutritional status and metabolic health in SIDS

Methods All nutrition focused interventions conducted within SIDS with change in nutrition or metabolic status as an outcome were eligible for inclusion. PRISMA framework was applied and MEDLINE, Embase, CINAHL, OARE library, Web of Science, SCOPUS, ASSIA, Econlit, AGORA, AGRICOLA, AGRIS, WHO-EMRO, LILACS were searched (2000 – 2021). Screening and data extraction was undertaken in duplicate; risk of bias was assessed using appropriate Cochrane tools. The results were narratively synthesised. This study was registered with PROSPERO (CRD42021236396).

Results From 44,699 unique records, 51 were eligible for inclusion. A total of 24 were found within the Caribbean, 8 within the Pacific and 19 within Singapore, Mauritius, and Seychelles. Most interventions were targeted at the individual level with community and healthcare being predominant settings. Twenty-nine studies were of a randomised study design and 22 of a non-randomised design. Most interventions (n=21) were embedded in a multidimensional approach, with others being limited to specific foods (n=6), supplements or fortified foods (n=11), nutrition education (n=6), dietary change (n=6) and policy (n=1). Interventions focused solely on dietary change including increased fruit and vegetable consumption or decreased consumption of foods high in saturated fat were least effective, those related to supplements, specific or fortified foods, and nutrition education was somewhat effective, while those taking a multidimensional approach most effective on nutrition or metabolic outcomes including Vitamin D status, BMI, HbA1C, lipids and weight change. These interventions included aspects of nutrition education, combining dietary change, supplements, and non-nutritional components including physical activity, behaviour therapy and self-medication management. Twelve studies had low risk of bias, 30 had moderate, and 9 high risk of bias.

Conclusion Of reported interventions, those with the most effective outcomes were embedded in a multidimensional approach. Our findings suggest that the impact of most interventions on nutritional or metabolic health have been evaluated at individual level. With the current rise of nutrition related public health challenges, there is need for further development and evaluation of these interventions at the population level.