Elaine Holmes

Background: Epidemiological evidence shows that fruit and vegetable (FV) intake reduces cardiovascular risk. Comparative effects of diets with different FV types on cardiometabolic disease risk markers remain unclear.

Objectives: This study aimed to compare the effects of standardised diets differing in FV types on vascular function, other cardiometabolic disease risk markers, urinary and plasma biomarkers in free-living adults with untreated prehypertension.

Methods: In a 9-wk randomised, controlled, crossover trial at 2 centres, 39 adults consumed standardised, provided diets with either 8 daily portions of common FV (apple, banana, pear, bell pepper, carrot, tomato), 8 daily portions of citrus and cruciferous FV, or 2 daily portions (low FV diet, control) for 2 wk per arm, with a 1-wk washout. Adherence was assessed using 24-h urinary potassium, sodium, and targeted plasma carotenoids and metabolites. The primary outcome was office blood pressure; secondary outcomes included pulse wave velocity (pwv), augmentation index standardised at 75 beats per minute (AIX75), lipids, and C-reactive protein (CRP). Between-group differences were assessed using linear mixed models with diet and period as fixed effects, and participant as random effect.

Results: Thirty-six participants completed the study (67% male, mean age 54±6 y; systolic BP 131.7±9.0 mmHg, total cholesterol: 5.4±1.1 mmol/L). No between-group differences were observed in office BP, PWV, AIX75, or CRP. Both common and citrus fruits and cruciferous FV diets tended to increase 24-h urinary potassium (by 6.94 mmol/24-h, P≤0.1; 8.0 mmol/24-h, P≤0.06, respectively), while 24-h urinary sodium remained comparable across diets (P≥0.87). Common FV diet significantly increased α- and β-carotene, and lycopene (P≤6.3x10-4), whereas citrus fruits and cruciferous FV increased lutein/zeaxanthin, β-cryptoxanthin, proline betaine, N-methylproline, and S-methyl-L-cysteine sulfoxide (P≤6.9x10-8). Common FV diet reduced total (-0.19 mmol/L, 95% confidence interval (CI): -0.32,-0.05), LDL (-0.15 mmol/L, 95% CI: -0.26,-0.03), and HDL (-0.05 mmol/L, 95% CI: -0.09,-0.00) cholesterol, whereas citrus and cruciferous FV reduced urinary creatinine (-1.19 mmol/24-h, 95% CI: -2.27,-0.11). No effects were observed on weight and physical activity.

Conclusions: Objective biomarkers confirmed FV adherence and suggest that FV types differentially modulate lipid and metabolic responses within 2 wk, without measurable vascular effects.

Introduction
The biomarkers and the mechanisms underlying the cognitive decline in ageing are not fully clarified although they might be effective for dementia prevention. We comprehensively explore metabolites associated with cognitive function by conducting metabolomics analysis in a cohort study of general Japanese men.

Methods
The study population was 672 Japanese men aged 46-83 years without stroke, who were randomly selected for the SESSA study from Kusatsu City, Japan. Metabolomics using liquid chromatography mass spectrometry (LC-MS) was conducted on plasma samples, focusing on lipid metabolism. The 442 metabolites were identified. The cognitive function was assessed by Cognitive Abilities Screening Instrument (CASI), and CASI score less than 82 was defined as having mild cognitive impairment (MCI). We examined the metabolites among groups with and without MCI using principal component analysis and partial least squares regression (PLS) or PLS rank order of groups. Multivariable adjusted logistic regression estimated the odds ratios (ORs) and 95%CI of MCI for the extracted metabolites per standard deviation. The 9 domain specific MCI using median value of each domain specific CASI score as cutoff points were also examined. KEGG pathway analysis was done to clarify the linked pathway for the extracted metabolites.

Results
The prevalence of MCI was 5.5%. The 22 metabolites were statistically significantly associated with CASI score (q <0.05 in PLS, 21 metabolites in the positive direction and 1 metabolite in the negative direction). Phosphatidylcholines (PC) and Lysophosphatidylcholines (LPC) were associated with low risk of MCI (OR(95%CI): PC16:0/20:3; 0.51(0.32-0.51), LPC18:2/0.0; 0.59 (0.36-0.96), LPC20:3/0.0; 0.48(0.28-0.82)), phenacetylcarnitine was associated with the high risk of MCI (OR (95%CI); 1.33 (1.01-1.75)) (q <0.05). The analysis for domain specific MCI extracted the metabolites for short term memory, language abilities and category fluency. The Linoleic acid metabolism pathway was statistically significantly associated with CASI score.

Conclusion
Glycerophosphocolines such as PC16:0/20:3, LPC18:2/0.0 and LPC20:3/0.0 were reduced in participants with MCI, and those metabolites associated with low risk of MCI. Linoleic acid metabolism might be important for prevention of dementia among Japanese.

Diet is a key modifiable factor for improving suboptimal lipoprotein profiles and reducing cardiovascular disease (CVD) risk(1). Dietary patterns like the Dietary Approaches to Stop Hypertension (DASH) or the Mediterranean Diet, with varying macronutrient components, have shown positive effects on total cholesterol and low-density lipoproteins (LDL)(2). However, limited research exists on the impact of different healthy diets on lipoprotein subclass profiles, which are increasingly known to influence CVD risk. This study aims to compare the nuclear magnetic resonance (NMR)-measured 112 lipoprotein profiles across three healthy dietary patterns: a carbohydrate-rich diet (CARB), similar to the DASH diet; a protein-rich diet (PROT); and an unsaturated fat-rich diet (USFA), similar to the Mediterranean diet. Lipoprotein parameters were generated using the Bruker IVDr Lipoprotein Subclass Analysis (B.I.LISA) method(3). The lipoprotein subclasses included different molecular components of very low-density lipoprotein (VLDL, 0.950–1.006 kg/L), low-density lipoprotein (LDL, density 1.09–1.63 kg/L), intermediate-density lipoprotein (IDL, density 1.006–1.019 kg/L), and high-density lipoprotein (HDL, density 1.063–1.210 kg/L). The LDL subfraction was further divided into six density classes, and the HDL subfractions were divided into four different density classes. Plasma samples from a randomised cross-over intervention study involving 156 individuals who completed more than two dietary patterns were included for the NMR analysis (registered at www.clinicaltrials.gov as NCT00051350 and NCT03369535). The Friedman’s test with post-hoc analysis, corrected for multiple testing, showed that all healthy dietary patterns led to a reduction in overall lipoprotein subclasses known to be associated with atherogenic risk. This reduction included large and medium-sized LDL subclasses, all intermediate-density IDL subclasses, as well as total plasma cholesterol, triglycerides, apolipoprotein-B100, apo-B100/apo-A1 ratio, and LDL-cholesterol (p < 0.05). Additional variations in lipoprotein subclasses specific to each diet were also observed. The PROT diet showed a decrease in small-sized and dense LDL, large to medium VLDL subclasses, and large-sized HDL subclasses. Conversely, the CARB diet exhibited an increase in smaller-sized and denser LDL, along with a decrease in large-sized HDL and an increase in smaller-sized HDL subclasses. The USFA diet led to decreases in LDL and overall VLDL subclasses, while increasing LDL and HDL subclasses (p < 0.05). The impact of different healthy diets with differential effects on lipoproteins suggests the possibility of targeting the cholesterol status of individuals to optimise lipoprotein profiles and thereby reduce CVD risk. Preliminary exploratory analyses based on linear mixed-effect models coupled with a latent profile analysis, adjusted for cholesterol status, showed that individual lipoprotein responses to specific diets varied. Inter-individual variations in lipoprotein responses to healthy diets were evident. A small proportion of individuals only responded to specific diets, suggesting potential of personalised nutrition based on individual lipoprotein profiles. These observed variations highlight the complexity of individual responses to dietary interventions.

Current nutritional rating systems, like the health star rating, help consumers understand the nutritional value of food and were designed in an effort to combat obesity. However, these systems have limitations, especially for edible oils, which vary widely in composition(1). Coupled with the lack of standardisation in ranking edible oils, there has been advocacy for the introduction of different nutritional scores for edible oils. This study aims to develop a simple and easy-to-use nutritional scoring index based on the composition of extra virgin olive oil (EVOO). The composition includes all fatty acid parameters and total polyphenol content, measurable by nuclear magnetic resonance (NMR) spectroscopy, thereby avoiding the need for multiple analytical platforms. The development of an EVOO nutritional score involved: i) establishing a unique consensus dietary reference index (DRIs) for each component and evaluating their impact on human health(2,3); and ii) computing Scoring Reference Values (SRVs) for each component, expressed as grams of component per 100 g of EVOO, based on the assumptions of a daily energy intake of 2000 kcal, with a fat intake of 35% of total caloric intake(2,3), and considering EVOO as the only source of fat. A nutritional score (0–100) was developed based on saturated fat (SFA), trans-unsaturated fat, oleic, linoleic, alpha-linolenic acids, and polyphenols. Components with more substantial effects/evidence on human health were given greater weight in the scoring. The developed index was subsequently applied to evaluate 314 EVOOs that passed the International Olive Oil Council (IOC) quality criteria. These oils were sourced from Australia (n = 94), Greece (n = 54), Italy (n = 54), Spain (n = 69), and Tunisia (n = 43) and analysed using 400 MHz NMR spectroscopy. Nutritional scores for all samples showed a mean of 62.3 (range 13 to 94), with Australian EVOOs exhibiting the highest mean score of 65, followed by Spain, Tunisia, Italy, and Greece. EVOOs were differentiated by their SFA content and the balance between polyunsaturated (PUFA) and monounsaturated fatty acids (MUFA). MUFA and PUFA were typically inversely related, except for two Australian oils that achieved high levels of both. This novel scoring index for EVOOs, grounded in health-related compositional parameters, facilitates the differentiation of EVOOs based on their nutritional value. This enables consumers to make informed choices regarding their oil selection. Given the rising prevalence of obesity and its associated morbidity, this tool is particularly significant. Additionally, the implementation of this nutritional index encourages producers to produce oils with superior nutritional profiles.

Understanding individual variability to dietary interventions is emerging as an important consideration in dietary interventions. Prior research has demonstrated ranging success of interventions. For example, Gardner and colleagues (2007) compared 4 weight loss diets in participants over a 12-month period(1), noting the range of weight loss was between 3.1kg to 6.3kg depended on individual and diet. Song and colleagues (2023) examined post prandial glucose response (PPGR) to four different carbohydrate meals. Dependent on the meal, the PPGR varied significantly between individuals(2). As such, it is inappropriate to assume that there is one dietary pattern appropriate for all individuals. Understanding the driving factors behind individual variation to specific foods and dietary patterns will allow us to tailor interventions to create optimal health outcomes for each individual. The aim of our study is to examine individual responses to different diets promoted for health. In our study, we investigated the biological diversity in response to the same dietary inputs among 23 participants at risk of type 2 diabetes and chronic disease over a two-week period. All participants completed four days on three dietary interventions (Mediterranean, Australian and low carbohydrate diets). Urine, serum, plasma, and faecal samples were collected, alongside the use of continuous glucose monitoring data, to explore the metabolic and glycaemic responses. Our findings reveal significant individual differences in blood glucose levels and metabolic outcomes. When examining fasting blood glucose levels, the low carbohydrate and Australian diets were optimal for 8 participants each, while the Mediterranean diet was optimal for 7 participants. However, this did not always correlate with post prandial blood glucose level optimisation. While blood, urine and faecal samples are yet to be analysed, these are expected to provide further understanding of individual biological responses. These results underscore the limitations of a universal dietary approach for optimising glycaemic control and highlight the necessity of personalised dietary recommendations that consider individual metabolic profiles. Our study provides crucial insights for future advances in precision nutrition, suggesting that personalised nutrition plans could lead to more effective management and prevention of T2D.