Phenotypic factors influencing the variation in response of circulating cholesterol level to personalised dietary advice in the Food4Me study

Individual response to dietary interventions can be highly variable. The phenotypic characteristics of those who will respond positively to personalised dietary advice are largely unknown. The objective of this study was to compare the phenotypic profiles of differential responders to personalised dietary intervention, with a focus on total circulating cholesterol. Subjects from the Food4Me multi-centre study were classified as responders or non-responders to dietary advice on the basis of the change in cholesterol level from baseline to month 6, with lower and upper quartiles defined as responder and non-responder groups, respectively. There were no significant differences between demographic and anthropometric profiles of the groups. Furthermore, with the exception of alcohol, there was no significant difference in reported dietary intake, at baseline. However, there were marked differences in baseline fatty acid profiles. The responder group had significantly higher levels of stearic acid (18 : 0, P= 0.034) and lower levels of palmitic acid (16 : 0, P= 0.009). Total MUFA (P= 0.016) and total PUFA (P= 0.008) also differed between the groups. In a step-wise logistic regression model, age, baseline total cholesterol, glucose, five fatty acids and alcohol intakes were selected as factors that successfully discriminated responders from non-responders, with sensitivity of 82% and specificity of 83%. The successful delivery of personalised dietary advice may depend on our ability to identify phenotypes that are responsive. The results demonstrate the potential use of metabolic profiles in identifying response to an intervention and could play an important role in the development of precision nutrition