Reducing free sugar intakes: Evidence for effective dietary recommendations.

Abstract

Reducing free sugar intakes: Evidence for effective dietary recommendations Lucy Ruth Boxall, Bournemouth University Large scale epidemiological investigations and RCT trials since the 1990’s have shown that diet is a leading risk factor for many communicable and non-communicable diseases. Within this research, diets high in free sugar intakes have been linked to conditions such as poor oral health, cardiovascular disease, obesity, depression and anxiety. Agencies such as the World Health Organisation have advocated for the use of food-based dietary guidelines to aid against the effects of poor diets, including those high in free sugars. Since then, over 90 countries have released their own national dietary guidelines, with many monitoring their nation's diets through programmes such as the UK’s National Diet and Nutrition Survey. Analysis of national reports have highlighted poor adherence to national dietary recommendations. Additionally, assessments into the effectiveness of these dietary recommendations are poor, with most countries having no way to monitor the impact of recommendations. This thesis sought to assess the research question of; what is the impact of current reducing free sugar advice, and its component nutrient, food and swap level advice on reducing free sugar intakes? The primary outcome was percentage free sugar intakes (FS%) and adherence to recommendations at an endpoint of 12 weeks. Secondary outcomes were assessed for a range of demographic, behavioural and taste variables, with qualitative interviews undertaken to investigate barriers and facilitators to intervention success. Using a randomised controlled parallel-group trial, 242 adults (18-65 years) were randomised across four trial arms to receive nutrient-based (N) (n=61), nutrient- and food-based (NF) (n=60), nutrient-, food- and food-substitution-based recommendations (NFS) (n=63) or no recommendations regarding free sugar intake (control, n=58). Data were analysed with intention to treat protocol and multiple imputation used in the case of missing data. Endpoint attrition was 17.5%. In the primary analyses, multiple regression models significantly predicted endpoint free sugar percentage intakes (FS%)(F(7,234) = 8.86, p<0.001, R2=0.21. Significant predictors were recommendations received (B=-0.636, p=0.029), baseline %FS (B=0.377, p<0.001) and baseline bodyweight (B=-0.04, p=0.041). The mean %FS reduced in all intervention groups N, NF, NFS by 2.47%, 3.25%, 3.08% respectively, in comparison to no change in the control group (-1.18%). Endpoint bodyweight was significantly predicted, F(6,235) = 1404.355, p < 0.001, R2 = 0.97, adj R2 0.97 with only the baseline bodyweight (b =0.952, p =0.001) and age ( b=-0.032, p =0.048) variables adding statistically significantly to the prediction. There was a weak association between endpoint bodyweight and group (b = -0.359, p=0.062). The mean endpoint bodyweight reduced in all intervention groups N, NF, NFS by 0.72kg, 1.44kg, 1.11kg respectively, in comparison to no change in the control group (-0.17kg). In exploratory analyses, change in: FS%, bodyweight and waist circumference was significantly correlated with summative adherence scores. Linear regression analyses found summative adherence significantly predicted change in FS%. Framework analysis of participant interviews identified seven themes and fourteen subthemes for investigations of barriers and facilitators to recommendations which provided insight when considering quantitative findings. Primary results show that providing participants with N, NF or NFS guidelines reduced FS% at an endpoint of 12 weeks. This supports the use of national dietary guidelines for dietary change. The average reduction of ~3% intakes achieved the same level of reduction as multiple years of the ‘Sugar reduction programme’ which included the same supporting aims of aiding in the reductions of free sugar intakes for the UK public. It was encouraging that the most adherent individuals had the largest reduction in FS%, but disappointing that national goals of <5 FS% intakes remained unmet. As seen worldwide, it is likely that poor adherence to dietary guidelines will be a persistent factor irrespective of information type provided. A limitation within this research was the potential of underreporting of sugar intakes. Underreporting is common across dietary monitoring trials, however, is of higher risk in the reporting of foods high in sugar, fat or salt. A second limitation can be regarded as participants completing their own research on reducing intakes. This does emulate what individuals are likely to do in a real-world context and was mentioned in participant interviews however, it makes observing differences between different recommendations less likely. This trial supports the use of simple targeted interventions for gradual dietary and physical change, with adherence being vital for long-term benefits. Further research is needed to identify individual factors for more effective dietary advice to be delivered on broader national scale rather than a one-size-fits-all approach.