Objective: There is growing evidence that childhood malnutrition is associated with non-communicable diseases (NCD) in adulthood and that body composition mediates some of this association. This review aims to determine if childhood body composition can be used to predict later-life cardiometabolic NCD and which measures of body composition predicts future NCD.Design: Electronic databases were searched for articles where: children aged under 5 years had body composition measured; cardiometabolic health outcomes were measured a minimum of 10 years later.Setting: The databases Embase, Medline and Global Health were searched through July 2020.Participants: Children aged under 5 years with a follow-up of minimum 10 years.Results: Twenty-nine studies met the inclusion criteria. Though a poor proxy measure of body composition, body mass index (BMI) was commonly reported (n 28, 97 %). 25 % of these studies included an additional measure (ponderal index or skinfold thickness). Few studies adjusted for current body size (n 11, 39 %).Conclusions: Many studies reported that low infant BMI and high childhood BMI were associated with an increased risk of NCD-related outcomes in later life but no conclusions can be made about the exact timing of child malnutrition and consequent impact on NCD. Because studies focussed on BMI rather than direct measures of body composition, nothing can be said about which measures of body composition in childhood are most useful. Future research on child nutrition and long-term outcomes is urgently needed and should include validated body composition assessments as well as standard anthropometric and BMI measurements.

There is growing evidence that childhood malnutrition is associated with non-communicable diseases (NCDs) in adulthood and that body composition mediates some of this association. This review aims to determine: if childhood body composition can be used to predict later-life cardiometabolic NCDs; which measures of body composition best predict future NCDs. Three electronic databases were searched for studies where: children aged under 5 year had body composition measured; cardiometabolic health outcomes were measured a minimum of 10 years later. 29 studies met the inclusion criteria. Though a poor proxy measure of body composition, Body mass index (BMI) was commonly reported ( n =28, 97%). 25% of these studies included an additional measure (Ponderal Index or skinfold thickness). Only some studies adjusted for current body size ( n =11, 39%). Many studies reported that low infant BMI and high childhood BMI increased the risk of NCD-related outcomes in later life but no conclusions can be made about exact timing of child malnutrition and consequent impact on NCD. Because studies focused on BMI rather than direct measures of body composition, nothing can be said about which measures of body composition in childhood are most useful. Future research on child nutrition and long-term outcomes is urgently needed and should include validated body composition assessments as well as standard anthropometric and BMI measurements.