Data presented in this paper were obtained from 52 clinically healthy participants at low-moderate risk (39 women; 13 men, with a mean age 49.1 ± 12.4 years) enrolled within the CARDIVEG (Cardiovascular Prevention with Vegetarian Diet) study, a trial conducted with the aim of comparing the effects of VD and MD on several cardiovascular disease (CVD) risk factors [10]. The study protocol and the general characteristics of the participants were extensively described in [13] and briefly summarized here. Participants with a low-to-moderate cardiovascular risk profile (< 5% at 10 years, according to the guidelines for CVD prevention of the European Society of Cardiology) [14] were recruited from the Clinical Nutrition Unit of Careggi University Hospital, Florence, Italy. Eligibility criteria included being overweight or obese (body mass index [BMI] ≥ 25 kg/m2) and the concomitant presence of at least one of the following CVD risk factors: total cholesterol levels > 190 mg/dL; LDL-cholesterol levels > 115 mg/dL; triglyceride levels > 150 mg/dL; fasting plasma glucose levels between 110 and 125 mg/dL. Ineligibility criteria were the presence of serious illness or unstable conditions, taking medication for any reason, being pregnant or lactating, the exclusion of meat, meat products, poultry, or fish from the diet in the past 6 months, or the participation in a weight loss program in the past 6 months.

The study was a randomised, open, crossover clinical trial with two dietary intervention periods, each lasting 3 months. The start of the study was preceded by a 2-week run-in period necessary to assess the motivation and availability of the participants and to obtain a 3-day dietary record (two weekdays and one weekend day), which was analysed by a dietician using a nutrition-specific database. The participants enrolled were randomly assigned to a MD (n = 27) or a VD (n = 25) group as a first dietary intervention and then crossed over to the other dietary treatment. All participants were instructed not to alter their lifestyle or physical activity grade during the study, and no weight loss goal was given. Written informed consent was obtained from each participant. Clinical evaluations were performed at the baseline before the start of treatment, 3 months after the start of the first dietary intervention (at the time of crossing over) and then 3 months after the start of the second dietary intervention. The primary outcomes of the study were differences in changes in body weight, BMI, and fat mass from the baseline. The secondary outcomes were differences in changes on circulating cardiovascular risk markers and apolipoprotein levels from the baseline.

The study was approved by the Ethics Committee of the Tuscany Region, Careggi University Hospital (SPE 15.054) and registered at clinicaltrials.gov (identifier: NCT02641834), and adhered to the principles of the Declaration of Helsinki and the Data Protection Act.

VD and MD were isocaloric between them, but hypocaloric with respect to the energy requirements of the participants. Both diets consisted of approximately 50–55% of energy from carbohydrate, 15–20% from protein and 25–30% from total fat (≤ 7% of energy from saturated fat, < 300 mg of cholesterol). The VD consisted of plant-based foods, eggs, milk and dairy products but completely excluded meat and meat products, poultry, fish and seafood. The MD was characterized by the consumption of all the food groups, including meat and meat products, poultry, and fish, although red meat consumption was limited to once a week. All participants were provided with a detailed 1-week menu plan, precise information on the foods that could be included or excluded, and a lot of different recipes for meal preparation. The dietary profiles for both VD and MD were calculated based on the portion sizes recommended by the Italian Recommended Dietary Allowances [15]. There was no difference between the two diets in the frequency of weekly servings of fruit and vegetables, cereals and olive oil. However, a higher frequency of weekly consumption of dairy products (21.5 vs. 18.5 servings), legumes (5 vs. 2.5), eggs (2 vs. 1), and nuts (2 vs. 1) was present in VD than in MD.

The adherence to the VD was assessed using a modified version of the National Health and Nutrition Examination Survey food questionnaire [16] and through unannounced telephone calls to participants, during which a 24-h diet recall interview was conducted. Participants were considered adherent to the VD if they reported no consumption of any animal flesh both in the questionnaire and in the interview. The adherence to MD was evaluated through the Medi-Lite validated questionnaire [17], considering participants who reported ≥ 10 points (in a scale ranging from 0 to 18) as adherents.

Data collection was carried out at the Clinical Nutrition Unit of Careggi University Hospital by standardised methods. All participants, who were required not to undertake strenuous physical activity during the day before the visit, were interviewed and examined between 6:30 a.m. and 9:30 a.m. after an overnight fasting period. Detailed information on demographics, risk factors, comorbidities, dietary and lifestyle habits was collected from each participant at the baseline. BMI, body composition and blood samples were obtained both at the beginning and at the end of the VD and MD intervention period. BMI was calculated as the weight (kg)/height (m2), with weight and height measured using a stadiometer. Body composition was determined by a bioelectrical impedance analysis device (TANITA, Arlington Heights, IL, model TBF-410).

Peripheral venous blood samples were collected at the beginning and at end of both dietary intervention periods into evacuated plastic tubes (Vacutainer; Becton Dickinson, Plymouth, UK). Samples were centrifuged at 4,000 rpm for 15 min (4 °C), and then stored in aliquots at − 80 °C. Circulating levels of apolipoproteins (ApoA-I, ApoB-100, ApoC-I, ApoC-III, ApoD, ApoE) were measured by Multiplex Luminex Assay using a custom kit (Bio-Plex Pro™ Human Apolipoprotein 10-Plex Assay, Bio-Rad), performed according to the manufacturer’s recommended protocol. Pro-and anti-inflammatory cytokines (interleukin (IL)-1ra, IL-4, IL-6, IL-8, IL-10, IL-12, IL-17, MCP-1, MIP-1β, VEGF, TNF-α, IP-10, IFN-γ) were determined by a Bio-Plex cytokine assay (Bio-Rad Laboratories Inc) according to the manufacturer’s instructions.

Statistical analysis presented in this paper was performed using the statistical package IBM® SPSS® Statistics for Macintosh, version 28.0 (IBM Corp., Armonk, N.Y., USA). The results were reported as means ± standard deviations (SD), frequencies and percentages, or geometric means and 95% confidence intervals (CI), as appropriate. All data were treated as paired samples from a crossover study. MD and VD interventions were analysed combining the results obtained in the two intervention periods of both groups. Differences between the two groups at the baseline were evaluated using the Mann–Whitney U test. The χ2 test was used for dichotomous variables. A Shapiro–Wilk test was performed in order to check if data were normally distributed. To evaluate the effects of the VD and MD a general linear model for repeated measurements, adjusted for order of treatment and weight change was conducted. Because these analyses assume normal data distribution, non-distributed data were logarithmically transformed and further analyses were carried out with the processed data. However, to facilitate interpretation, the log data were reconverted to the original antilogarithmic scale and presented here as geometric means and 95% confidence intervals (CI). The possible dietary carryover effect, that is, the effect that considers whether the impact of the first treatment is still present when the patient enters the second treatment period, was analyzed. We evaluated the sequence effect, which considers whether the impact of VD and MD was different when the order of administration changed. This effect was estimated by comparing the geometric mean change difference between treatments in the VD group and in the MD group, after adjustment for order of treatment. Subgroup analyses were performed to evaluate possible differences in the changes of circulating apolipoprotein levels according to some characteristics of the study population, such as sex (women and men), age (less than or equal to 50 year-old and older than 50 years-old) or cardiovascular risk factors (total cholesterol levels > 190 mg/dL; LDL-cholesterol levels > 115 mg/dL; triglyceride levels > 150 mg/dL; fasting plasma glucose levels between 110 and 125 mg/dL). A Spearman’s correlation analysis was conducted to evaluate the relationship between changes in circulating apolipoprotein levels and changes in circulating lipid profile, inflammatory profile and dietary composition. Finally, a linear regression analysis adjusted for order of treatment and weight change was conducted in order to evaluate the relationship between changes in circulating apolipoprotein levels and changes in circulating lipid profile, inflammatory profile and dietary composition. P values < 0.05 were considered statistically significant.