The main aim of this study is to investigate the relationship between the healthy eating index-2015 (HEI-2015) and food security with CVD risk factors in type 2 diabetes patients in a cross-sectional design. Individuals with type 2 diabetes were recruited from the Tehran University of Medical Sciences affiliated Subspecialty clinic of diabetes and metabolic diseases. Collecting the data was done between May 2021 and September 2022. The participants in this research were 490 randomly chosen people with aged 35–80 referred to a Subspecialty clinic for diabetes and metabolic diseases.

To be eligible, participants needed to meet the following conditions: have type 2 diabetes for at least 2 years, be between the ages of 35 and 80, not be on insulin, not be pregnant or breastfeeding, not be receiving estrogen hormone therapy, and not have an autoimmune disease, acute gastrointestinal disorder, acute renal disease, or liver cancer. Exclusion criteria included a lack of information in the medical records and reporting of dietary energy intake (<800 kcal/day and over 4,200 kcal/day). The research protocol was accepted by the Human Ethical Committee of Tehran University of Medical Science [IR.TUMS.MEDICINE.REC.1400.185].

The sample size for this research was chosen by the prevalence of abdominal obesity in the nation in proportion to the quality of the overall diet as indicated by the index (HEI) (27). Using the formulas P1: the prevalence of abdominal obesity in persons with poor diet quality (0.31), P2: the prevalence of abdominal obesity in individuals with a high diet quality (0.39), C 1 − B  = − 0.842, and c a / 2  = 1.96, sample size was determined:

Q1: 1 − P1 = 1 – 0/31 = 0/69 n1 = m.

Q2: 1 − P2 = 1 – 0/39 = 0/61 n2 = m × r.

n₁ = m, n₂ = m × r

n₁ + n₂ = 345 + 145 = 490

According to this formula, 490 people were needed.

Before the commencement of the study, participants completed written permission forms confirming that they understood and agreed to participate. Those who did not provide informed consent were not included in the study. Patients referred to the clinic must undergo a new laboratory test. As a result, biochemical indicators such as Fasting Blood Sugar (FBS) and LDL-c were obtained from their medical laboratory test. Personal information like age, drug history, taking supplements, and other data were collected in interviews with patients by a dietitian.

The socioeconomic status demographic questionnaire, which contained questions about marital status, education, employment, family size, means of support, method of transportation, having a private house, etc., was utilized for this purpose. In order to create the socioeconomic status score, the codes were attached to each questionnaire item. After that, the mean score was reported from 10 to show socioeconomic status score.

An accurate and valid 168-item semi-quantitative food frequency questionnaire (FFQ) was used to estimate typical dietary consumption (28). An experienced nutritionist gathered nutritional data via a face-to-face interview. We calculated the daily intake of each food item and then converted them to grams per day. Then, the Nutritionist IV program was used to calculate macro- and micronutrient intakes (First Databank Division, the Hearst Corporation, San Bruno, CA, United States, modified for Iranian foods).

The 2015 HEI contained nine sufficiency components (total fruit, whole fruit, total vegetables, greens and legumes, whole grains, dairy, total protein sources, seafood, plant protein, and fatty acids) and four moderate components (refined grains, sodium, percentage of energy from added sugars, and percentage of energy from saturated fatty acids). Except for the unsaturated to saturated fatty acid ratio, each component is assessed on a density basis out of 1,000 calories. Because several surveys lacked brand specificity and information regarding discretionary salt usage, we changed the sodium component score. Total fruit has 100% fruit juice, while whole fruit includes all forms except juice. Green beans contain legumes, and dairy products include milk. All milk products, such as fluid milk, yogurt, and cheese, as well as fortified soy drinks, are included. Seafood, nuts, seeds, soy products (other than drinks), and legumes are examples of marine food and plant proteins (beans and peas). Based on the distribution of reported salt consumption (mg/d), we classified the individuals into 11 equal groups and awarded corresponding scores ranging from 0 to 10 (higher score for less sodium consumed). Table 1 in the appendix shows the specifics of the HEI-2015 scoring standards. The overall HEI-2015 score ranges from 0 (no adherence) to 100 (complete adherence) (29).

In a fasting condition, common laboratory procedures were used to measure fasting blood sugar (FBS), total cholesterol, high-density lipoprotein cholesterol (HDL-c), low-density lipoprotein cholesterol (LDL-c), and triglycerides (TG) (after 12–14 h overnight fasting). In a seated posture, a physician measured the systolic (SBP) and diastolic (DBP) blood pressures using a sphygmomanometer.

By using LDL-C, HDL-C, and TG and following formula the novel risk factor for cardiovascular disease like: CRI-1,CRI-2, LAP, AIP, and CI were obtained:

Patients were weighed in light clothes and barefoot on a digital scale (SECA, Hamburg, Germany) to the nearest 0.1 kg. Standing height was measured using a stadiometer set on the wall, with accuracy to within 0.5 cm. Using a measuring tape, determine the waist circumference (WC) of the patient at the narrowest point, immediately above the belly button. A dietician measured the hip circumferences of patients by placing a tape measure around the widest part of the hips.

The following formulae were used to determine body mass index (BMI):

Body mass index is calculated as kg/m². ABSI and BRI and abdominal volume index (AVI) were computed using previously described methods using WC (m), BMI (kg/m²), and height (m) as below:

We utilized the International Physical Activity Questionnaire (IPAQ) short form to measure physical activity levels (30). The IPAQ consisted of seven questions. Overall, the questions assessed the number of days and minutes spent participating in light and heavy activities, as well as the average time spent walking and sitting during the previous 7 days. Physical activity levels of 600 METs-min/week were classed as low, 600–3,000 METs-min/week as moderate, and > 3,000 METs-min/week as high.

We considered Five-point MACE for this study. Myocardial infarction, stroke, hospitalization for heart failure, and revascularization treatments, including angioplasty and bypass surgery, are all included in the definition of Major Adverse Cardiac Events (MACE). Since we do not have CV death for our study population, we put it aside.

Patients were asked about their history of chronic diseases. Chronic disease history includes Thyroid, blood pressure, cardiovascular disease, chronic kidney disease, and chronic obstructive pulmonary disease.

First of all, subjects were classified based on tertiles of HEI-2015 scores (T1: 43–59; T2: 60–66; T3: 67–88). Baseline quantitative and qualitative variables were compared across HEI tertiles by Analysis of Variance (ANOVA) and Chi-square tests, respectively. Continuous variables were reported as mean (SD), and categorical variables were shown as numbers and percentages. Dietary intakes of HEI components and other nutrients were evaluated across HEI tertiles using Analysis of Covariance (ANCOVA). In this regard, total energy and macronutrient intakes were adjusted for age and gender, but other nutrients were additionally controlled for energy (Kcal). The associations of HEI with CVD risk factors and anthropometric indices were explored through binary logistic regression. Based on previous investigations (31, 32), we considered ABSI, BRI, AVI, and AIP cut-off points to be 0.08, 5.20, 17.30, and 0.11, respectively. Values above the mentioned cut points were considered as the dependent variable in this study. For other CVD risk factors (CRI-1, CRI-2, Cholindex, and LAP), participants were categorized into tertiles of each variable. As there is no specific cut-off point for these variables, the last tertile of each risk factor was considered the dependent variable in the logistic regression. Odds ratios (ORs) and 95% confidence intervals (95% CIs) were reported for the association of HEI with CVD risk factors and anthropometric indices in crude and adjusted models. In case of anthropometric indices, two multivariable-adjusted models were applied. For the first model, we controlled the confounding effects of age, gender. and energy intake. The second model was further adjusted for smoking, socioeconomic status, chronic disease history, MACE history, diabetic diets, education level, marital status, physical activity, and supplement use. Considering CVD risk factors, we also included blood pressure and heart disease medications in the second model.

Furthermore, an additional model (model 3) was considered for CVD risk factors in order to control the confounding effect of BMI as well. The first tertile of HEI was considered the reference for all analyses. The trend of ORs across tertiles of HEI was obtained by considering HEI tertiles as an ordinal variable in the logistic regression models. All statistical analyses were performed by SPSS software (version 26; SPSS Inc., Chicago, IL). p < 0.05 was considered a significant level.

The study included 490 adults in total. Table 1 displays patient socio-demographic characteristics based on HEI-2015 tertiles. The mean age of the patients across the tertiles was 61.19, 62.94, and 64 years (p = 0.03). The tertiles 1 group has the highest BMI, with a mean of 28.09 kg/m². Most variables had no significant differences between tertiles except for hip circumference (p = 0.02) and supplement user (p = 0.005). Of 490 individuals, 66.8% of patients used blood pressure medication, and just 37.1% followed a diabetic diet.

Table 2 describes patients’ biochemical, novel CVD risk factors, and anthropometric indices. SBP and DBP averaged 125.44 ± 15.17 mmHg and 77.89 ± 7.88 mmHg, respectively. The first tertile had the highest FBS and HbA1C, with mean values of 157.53 mg/dL and 8.17%, respectively. There are significant differences in BRI (p = 0.04) but only marginally in LAP (p = 0.07).

Significant differences (p < 0.001) were observed in the weighted proportions of participants who received the highest component score on the HEI-2015 component. There was also identification of the participant who received the least sodium (p = 0.05; Table 3).

Participants in the highest tertile of the HEI-2015 consumed more protein (p = 0.001), calcium, magnesium, fiber, vitamin B2, and vitamin B9 (p < 0.001). There are no significant differences in the intake of certain nutrients, such as selenium (p = 0.47), vitamin B6 (p = 0.55), and iron (p = 0.61; Table 3).

In the crude model and after adjusting for potential confounders such as age, energy, smoking, welfare, CDH, diabetic diet, BP medication, and heart medication, patients who had more adherence to HEI had lower AIP (OR: 0.56; 95% CI: 0.34–0.94; p-trend =0.02) (Table 4).

For CRI-I, participants with more adherence to HEI had a lower odds ratio of CRI-I (OR: 0.61; 95% CI: 0.38–1.00; p-trend = 0.05), although this signification disappeared after the adjustment.

Participants who had higher adherence to HEI-2015 were 59% fewer odds of BRI (95% CI: 0.37–0.94; p-trend = 0.02), and this association was still significant after adjusting for all the confounders (OR: 0.52; 95% CI: 0.29–0.95; p-trend = 0.03). There was no signification in other anthropometric indices like; ABSI and AVI (Table 5).