Examining the Effect of a High Quality Dietary Intervention on Cognitive Function in Early Adolescence

Introduction: Excessive consumption of high fat, high sugar foods may precipitate cognitive decline. This effect may be more pronounced during cognitive development. The present single-blind randomized controlled trial (RCT) was conducted to examine the effect of a moderate fat, low added sugar (MF/LS) dietary intervention on cognitive function in 8-11 yr old preadolescents with a pre-established high fat, high sugar (HF/HS) dietary pattern. Participants included 17 non-obese (BMI Percentile: 25.4 - 91.3) low to middle income preadolescents randomized to 2 weeks of their usual HF/HS diet (control) or a MF/LS intervention diet. Method: The MF/LS intervention diet was restricted to 25% of calories/day from total fat and <10% of calories from added sugar while the HF/HS control diet was maintained at > 40% of calories/day from total fat and >15.9% of calories from added sugar. All food served was measured to the tenth of a gram. Any uneaten portion of food was weighed to obtain accurate measures of actual intake. NDSR dietary analysis software was used to assess macronutrient, micronutrient and added sugar intakes. Participants were weighed weekly to ensure they remained in energy balance throughout the duration of the study. Pre-post cognitive assessment served as the primary outcome measure. A battery of age appropriate tests from the Penn Computerized Neurocognitive Battery (CNB) as well as the widely used and a previously validated Trail Making task were selected to assess executive function, speed of processing, working memory, attention and spatial ability. Results: Both ANCOVA and a repeated measures approach were used to evaluate the mean difference of post-intervention scores between conditions, controlling for pre-intervention scores and other covariates including age, gender, sleep and mood. For each statistical approach, 10 tests were run, encompassing each of the cognitive assessments given and, for some, their delayed counterpart. Based on the ANCOVA analysis, participants randomized to the MF/LS intervention had a faster median response time (RT) for correct responses on 2 of the 10 tests analyzed, including the initial facial recognition task and its delayed counterpart. Compared to controls, the intervention group displayed 1) a faster total correct RT while controlling for gender (p = 0.02), 2) a faster true negative RT when controlling for gender and age (p = 0.012), and 3) a faster delayed task median total correct RT when controlling for gender and age (p = 0.005). No significant differences between groups were detected for the other assessments. Based on a repeated measures approach, none of the 10 tests analyzed reached statistical significance. Multiple regression analyses revealed a dose response effect on face recognition RT based on % intake of daily calories from total sugar, added sugar, total fat and saturated fat such that a 10% increase in % calories from total sugar, added sugar and saturated fat decreased processing speed for total correct responses on the initial facial recognition task by 0.58 seconds whereas a 10% increase in % total fat decreased processing speed on the same task by 0.44 seconds. The multivariate regression analyses controlled for gender and pretest scores. Conclusions: A 2-week MF/LS dietary intervention may improve delayed face recognition in low to middle income preadolescents with a pre-established HF/HS dietary pattern. Although the intervention appeared to demonstrate a positive effect on 2 measures of cognitive function (initial and delayed facial recognition), after Bonferroni correction, these results only remained significant for the delayed task median total correct RT when controlling for gender and age (p = 0.005). Thus, study results must be interpreted with caution as they may simply be an artifact of chance finding in the ANCOVA statistical analysis. Further investigation of benefits proffered by decreasing % total sugar, % added sugar, % total fat and % saturated fat intake to preadolescent cognition is warranted. Future work should focus on replicating the present study in a larger sample, using hippocampal-dependent specific tasks.