Patients and study design

We carried out a multicenter, open-label randomized controlled pilot trial, according to Helsinki Declaration (IV Adaptation) and European Guidelines for Good Clinical Practice. The trial was designed in order to adhere to CONSORT guidelines. Being a pilot study, we established a minimum target of 12 participants for each group [22]. The study protocol was approved by the Ethical Committee of the Azienda Ospedaliero-Universitaria Consorziale Policlinico of Bari, Italy (Authorization nr. 1644/2014, 16st December 2014) and registered in the ClinicalTrials.gov registry database (registration nr. NCT03127553, on 25/04/2017). Hypertensive patients were enrolled according to the following inclusion/exclusion criteria: (i) Inclusion: arterial hypertension stage I-II (BP between 130/80 mmHg and 145/95 mmHg), aged between 50 and 80 years, stable control of BP, estimated glomerular filtration rate (eGFR) > 60 ml/min/1.73m [2], signed informed consent; (ii) Exclusion: diuretic therapy, need for more than three antihypertensive drugs, proteinuria > 1 g/day, celiac disease, systemic inflammatory diseases, suspect/clinical diagnosis of malignancy, chronic liver disease, corticosteroids/immunosuppressive therapy, previous acute CVDs (myocardial infarction, stroke), psychiatric conditions reducing the compliance to protocols. Primary outcome: reduction of sodium intake, measured as 24-h urinary sodium concentration. Secondary outcomes: reduction of BP, gut microbiota modulation, evaluation of endothelial stress through flow-mediated dilation (FMD). All the outcomes were evaluated after six months of intervention.

Volunteers were recruited in two different ways. The Bari Unit recruited patients amongst the hypertensive patients assisted by two general practitioners’ (GP) offices (Dr. Loiudice private office—Altamura Italy, and the associated general practitioner clinic “CPT Europa” – Bari Italy). The GPs cooperated to this study by informing the potentially eligible patients and by collecting an informal consent to be contacted by the medical personnel involved in the study for eligibility assessment. The Foggia Unit recruited patients attending the Unit of Hypertension of the Department of Surgical and Medical Sciences. The study was carried out at outpatient centers of Nephrology and Cardiovascular Disease Units, Department of Emergency and Organ Transplantation – University of Bari (n = 26 patients enrolled) and of Nephrology Unit, Department of Surgical and Medical Sciences – University of Foggia (N = 31 patients enrolled). The recruitment started in December 2014 and ended in June 2015. The study was planned to end before summer, in order to avoid important losses of sodium through sweating. Moreover, to counteract the shift of environmental temperature from December to June, all patients were enrolled in the same month and completed the study in the same month. In this way, seasonal variation of Na excretion are compensated by the parallel arm design of the study.

Randomization and nutritional counseling

At enrollment, each participant signed a written informed consent. Patients belonging to each participating center were randomly allocated into three study arms (A, B and C) in a 1:1:1 ratio by an independent researcher by using a software, with gender and smoking as blocking factors. In both centers, patients were enrolled by a physician and assigned to nutritional intervention by a dietitian. After a literature review and evaluation of feasibility of LS diet, also considering the salt content in experimental and standard bread [2, 4, 10], a sodium reduction target of 2300 mg/day was set.

Group A represented the control group, with patients undergoing a six-month free-diet (FD), including the use of the standard “Oropan” Altamura bread (containing 750 mg Na for 100 g product). Patients randomized in the group B followed a six-month LS diet, including the use of standard “Oropan” Altamura bread (containing 750 mg Na for 100 g), while patients randomized in the group C followed a six-month LS diet, including the use of the innovative LS “Oropan PANDELCUORE” Altamura bread (containing 280 mg Na for 100 g). Group B and C diets were normocaloric, normoproteic, LS (2300 mgNa/day), Mediterranean-style (60% carbohydrates, 15% proteins, 25% fats), including 5 daily servings of fruits and vegetables, 120 g of experimental (either standard or LS) bread, 1 daily serving of milk/yogurt, 1 daily serving of pasta/whole-grain cereals, 1 daily serving of a protein dish, according to the following weekly frequency: 2–3 times legumes, 1–2 times fish, 1–2 times meat (preferably white), 1 time eggs. Patients in both groups received differential indications about the allowed amount of table salt (respectively, 2.5 g and 4 g) in order to respect the total daily amount of salt prescribed. The nutritional composition of the two breads is reported in Supplementary Table 2. At the enrollment and throughout the study, each patient received, twice a week, a supply of either standard or LS bread, depending on the randomization, providing an amount of three bread slices (120 g) per day.

At the beginning (T0) and at the end (T6) of the study period, each patient underwent medical examination, including: resting office BP measurement, height and weight measurement, 12-h fasting venous blood and 24-h urine collection for determining urinary and blood chemistry profile, stool collection, FMD analysis. Moreover, they underwent a nutritional visit including anthropometric and food intake evaluation (by a food frequency questionnaire and 24 h recall). All the data were registered on an electronic case report form. No incentive was provided to the volunteers, while the bread was provided free of charge.

Blood and urine chemistry

Fasting morning blood samples were used to measure serum glucose using Siemens enzymatic method (Dimension Vista 1500, Siemens Health Diagnostics, Deerfield, IL), glycated haemoglobin (HbA1c) levels by high-performance liquid chromatography (BioRad D10, Pratteln, Switzerland) and creatinine by enzymatic method (Dimension Vista 1500, Siemens Health Diagnostics, Deerfield, IL). Additional aliquots of serum were obtained, after centrifugation, for each patient/time point and stored at − 80 °C until use. Moreover, 24-h urine samples were collected to measure the sodium concentration. Briefly, timed 24-h urine specimen collection was started in the morning before the visit, after discarding first morning urine sample. The entire volume of urine was collected in the clean, unused container with 2–3 L volume. Total volume of the collection was measured. According with the standard procedure, urinary sodium was measured using ion selecting electrode method by V-Lyte IMT autoanalyser (Dimension Vista® 1500, Siemens Healthcare Global, Erlangen, Germany).

Blood pressure and Flow-mediated dilation of brachial artery

Office BP was measured in the supine position using a sphygmomanometer with an appropriate cuff on the left arm; the average of two readings was used for statistical analysis. Hypertension was defined as BP > 130/80 mmHg or currently on antihypertensive treatment [23].

In the Bari unit subgroup (n = 26 patients), endothelium-dependent vasodilatation FMD of the brachial artery was non-invasively assessed, using high-resolution ultrasound in a quiet, air-conditioned environment (22–24 °C), on the same day of the visit and blood sampling. The patients fasted for at least 8–12 h. The study was performed using an image analysis system software, certified by the CNR of Pisa (MVE II), as previously described [21]. Briefly, subjects were placed in supine position, a sphygmomanometer cuff was inserted distally to the brachial artery and image acquisition was made before, during and after an ischemic stimulus. The FMD was calculated as described elsewhere [24]. The measures of FMD showed good reproducibility, with an ICC of 0.95.

Enumeration of cultivable bacteria

Fecal samples were collected at each time point by the Bari unit subgroup (n = 26 patients, group A n = 9, group B n = 8, group C n = 9). After collection, samples were immediately diluted with Amies Transport medium (Oxoid LTD, Basingstoke, Hampshire, UK) (ca. 15 g, 1:1 wt/wt), under anaerobic conditions (AnaeroGen, Oxoid LTD, Basingstoke, Hampshire, UK). Samples mixed with Amies Transport medium were immediately analysed (plate counts). Fecal samples (5 g) were mixed with 45 ml sterilized physiological solution and homogenized. Counts of viable bacterial cells were carried out as described elsewhere [25]. The following selective media were used: Wilkins-Chalgren anaerobe agar (total anaerobes); Plate count agar (total aerobes and anaerobes); MRS agar (Enterococcus, Lactobacillus and Leuconostoc); Slanetz and Bartley (Enterococcus); Rogosa agar, plus 1.32 ml/l of glacial acetic acid (Lactobacillus); M17 (Lactococcus and Streptococcus); Baird Parker (Staphylococcus); Reinforced Clostridial Medium supplemented with 8 mg/L novobiocin, 8 mg/L colistin (Clostridium); Wilkins-Chalgren anaerobe agar, plus GN selective supplements and sheep blood defibrinated (Bacteroides, Porphyromonas and Prevotella); MacConkey agar No2 (Enterobacteriaceae); Chromocult (total coliform) (Merk, Darmstadt, Germany, Europe); GSP agar (Sigma-Aldrich, St. Louis, MO), plus penicillin-G (60 g/l) (Pseudomonas, Aeromonas); Bifidobacterium agar modified (Bifidobacterium) (Becton Dickinson, Le Pont de Claix, SA, France). Except for Bifidobacterium agar modified, Chromocult and GSP agar, all media were purchased by Oxoid Ltd (Basingstoke, Hampshire, UK).

Community-level catabolic profiles

Carbon source utilization patterns of the fecal microbiota were assessed using Biolog 96-well Eco micro-plates (Biolog, Inc., Hayward, CA). Micro-plates contained a variety of carbon sources (carbohydrates, carboxylic acids, polymers, amino acids, amines, and miscellaneous substrates), and were assessed in triplicate. Five grams of feces mixed with Amies Transport medium (1:1) were treated as previously described [20]. The microbial suspension was diluted into sterile chloride solution and distributed into each a 96-wells Biolog Eco micro-plates. They were incubated at 30 °C in the dark on slow stirring, and a micro-plate reader (Biolog Microstation) was used to measure 590 nm absorbance every 24 h. Three indices were determined [26]. Shannon’s diversity (H’), substrate richness (S), and substrate evenness (E) were calculated, respectively indicating the substrate utilization pattern, the number of different substrates used and the equitability of activities across all utilized substrates. They were calculated as reported elsewhere [20].

Statistical analysis

24-h urinary sodium and systolic and diastolic BP were defined in terms of variation (T6-T0) and expressed as medians and interquartile range (IQR). Data distribution was assessed by D'Agostino&Pearson normality test. Normally distributed data are presented as mean ± standard deviation, non-normally distributed data as median and interquartile range. Differences between quantitative variables were analyzed by unpaired Student t-test (normally distributed data) or Mann–Whitney (non-normally distributed data) test as appropriate. Additionally, one-way ANCOVA was performed to compare the effectiveness of three nutritional interventions on sodium excretion, using baseline urinary sodium, systolic and diastolic BP as covariates. Levene’s test and normality checks were carried out and the assumptions met. p-values from microbial counts analysis were adjusted for multiple comparisons using the false discovery rate (FDR). p-values < 0.05 were considered significant. All the analyses were performed using SPSS software (version 21; IBM, Armonk, NY).