Fourteen (six males and eight females) apparently healthy, non-smoking (age 20–55 years) participants were recruited through university email distribution lists. Exclusion criteria included taking medication for heart disease, hypertension, high cholesterol or diabetes, and the use of antioxidant and vitamin supplements. The study was conducted in accordance with the guidelines laid down by the Declaration of Helsinki and approved by the Research Ethics Central Committee at Queen Margaret University; Edinburgh, UK. All participants gave written consent and completed a health status questionnaire to determine their eligibility.
The study was a 14-day small-scale intervention of a daily intake of PM versus a WP variety. Following a 7-day run-in period, eligible participants were randomly assigned to receive a 200 g/day serving of cooked PM or WP. All potatoes were from the same batch and cooked from fresh each day (unpeeled and boiled for 15 min). Participants followed each intervention arm for 14 days, after which they were crossed-over to the next arm of the study separated by a 7-day washout period. A list of forbidden phenolic-rich foods and beverages (including green tea, black tea, coffee, red wine, dark chocolate and berries) was provided and participants were also advised to limit fruit, vegetable and potato intake over the study period. Compliance was assessed by direct observation of consumption of the meal, which was provided as a daily lunch for each participant at Queen Margaret University for the duration of the study. Food records based on a 3-day diary were used to assess compliance before, during and after each arm of the intervention.
Purple Potato and White Potato Analysis
Potatoes from the same batch were supplied by Albert Bartlett Ltd., Airdrie, UK, and stored in the dark at 4 °C throughout the study period. The nutritional composition of the potatoes was provided by the supplier, with each 200 g serving of PM and WP containing; 154 kcal, 34.8 g carbohydrate, 4 g protein, 0.2 g fat, 2.1 g fibre and 1.6 g sugars. Potatoes were analysed prior to consumption to determine levels of total phenolics (TP), total anthocyanins (TA) and antioxidant capacity (AOX). Raw and cooked potatoes were diced and freeze-dried in vacuo overnight. Extracts were reconstituted in 50:50 methanol:water (v/v) and 10:90 acetone:water (v/v) and analysed for TP by the Folin–Ciocalteau method  and AOX capacity by the ferric-reducing antioxidant power assay , respectively. TA were determined using an adaptation of the pH shift method . Osprey was selected to act as a control WP because it matched the energy and nutritional composition of PM and contained negligible anthocyanins.
Liquid Chromatography Mass Spectrometry (LC-MS) Analysis
One hundred mg of freeze dried potato was extracted in 50% (v/v) aqueous acetonitrile (ACN) containing 0.1% (v/v) formic acid (FA) and rotated at 100 rpm for 1 h in the dark at 4 °C . Samples were centrifuged (5000 g for 15 min at 4 °C), dried and re-suspended in 500 mL of 5% ACN/0.1% FA. A LCQ-Deca LC-MS system, comprising an autosampler, pump, photodiode array detector (PDA) and an ion-trap mass spectrometer (ThermoFinnigan, UK) was used for sample analysis. The PDA scanned discrete channels at 280, 365 and 520 nm. Samples were applied to a C18 column (Synergi Hydro C18 with polar end-capping, 2.0 × 250 mm, Phenomenex, UK) and eluted using a linear gradient of 5% ACN/0.1% FA to 40% ACN/0.1% FA over 35 min at a rate of 200 ml min−1. Electrospray ionization analysed the samples in positive and negative ion modes in full scan analysis followed by data-dependent MS/MS of the most intense ions using collision energies of 45%. The capillary temperature was set at 250 °C, with sheath gas at 60 psi and auxiliary gas at 15 psi. Peaks were identified by comparing their relative retention times, PDA spectra, mass to charge ratios (m/z) and MS2 properties with previous reports [17, 18]. Components were quantified by their UV maxima peak areas calculated using the resident software and expressed as average ± standard errors (n = 3). This approach is not quantitative but gives valid relative comparisons of the components between different samples or treatments.
Participants attended the University at the start and at the end of each treatment arm for vascular measures, anthropometric measures and blood sampling. A SECA 709 mechanical column scale was used to measure body weight to the nearest 0.1 kg and height was measured to the nearest 0.1 cm with a SECA 220 telescopic measuring rod (SECA, Birmingham, UK). Fasted blood samples were obtained by venepuncture into EDTA and lithium heparin tubes. Plasma was immediately separated by low-speed centrifugation (2500 g for 10 min) and stored at -80C prior to analysis. Assessment of lipid profile; high-density lipoproteins (HDL), low-density lipoproteins (LDL), and triglycerides (TAG), in addition to c-reactive protein (CRP), insulin and glucose were undertaken at the Routine Clinical Biochemistry Laboratory, Western General hospital (Edinburgh, UK), using an automated platform (Olympus, UK). Homeostasis model assessment of insulin resistance (HOMA-IR) was calculated based on published equations . A validated automated A&D Medical UA-767 BP monitor (A&D medical, San Jose, CA, USA) was used to measure arterial blood pressure (BP) after a 10 min episode of horizontal resting on the subjects’ right arm using an appropriate BP cuff. Three readings were taken at 2 min intervals and mean systolic blood pressure (SBP) and diastolic blood pressure (DBP) were calculated from the second and third readings to increase the reliability of the results. Pulse wave velocity (PWV) was measured between the carotid and femoral artery (PWVcf) by means of a validated Vicorder™ device (Skidmore Medical Limited, Bristol, UK). This was performed by measuring the distance between the midpoints of two oscillometric cuffs, placed at the collar (carotid artery) and at the proximal right femur (femoral artery). A correction factor of 0.8 was used to account for the difference between the tape-measured distance between both cuffs and the reference distance, in accordance with the American Heart Association guidelines . The transit time, recorded automatically, reflected the time lag between pulse wave registration at the carotid and femoral cuffs, which subsequently was divided by the distance. The mean PWVcf of three measures was recorded and the data expressed in m/s.
All statistical analyses were conducted using SPSS version 22.0 (SPSS Inc., Chicago, IL, USA). Baseline characteristics are presented as mean values ± standard deviations. Variables were examined for normality and skewness using Shapiro-Wilko tests, and normality was assessed with the Shapiro-Wilko test and inspection of Q-Q plots. The effect of the dietary intervention on the outcome variables was analysed using generalised linear models with Bonferroni correction after adjustment for baseline values, sex and BMI. Interaction between treatment and baseline covariates were assessed and results presented for the total group. P-values were two-sided, and treatment effect was considered statistically significant at P ≤ 0.05. Analysis of the significance of any difference observed between raw and cooked PM and WP for TP, AOX capacity and TA analysis was carried out using a paired t-test, and differences were considered statistically significant at P ≤ 0.05.