Mediterranean diet, neutrophil count, and carotid intima-media thickness in secondary prevention: the CORDIOPREV study

Abstract

Background and Aims

Several studies have supported the role of innate immune system as a key factor in the sterile inflammation underlying the pathophysiology of atherosclerosis in mice. However, its involvement in humans remains unclear. This study aimed to explore the association between neutrophil count, and the intima-media thickness of common carotid arteries (IMT-CC), as well as the potential impact of long-term dietary interventions on these associations.

Methods

A comprehensive analysis was conducted within the framework of the CORDIOPREV study, a long-term secondary prevention study involving dietary interventions with either a Mediterranean or a low-fat diet. The study evaluated the relationship between absolute neutrophil count and neutrophil-related ratios with IMT-CC at baseline and after 5 and 7 years of dietary intervention.

Results

At baseline, patients in the highest tertile of neutrophil count had a higher IMT-CC and number of carotid plaques, when compared to lowest tertile (P < .01 and P < .05, respectively). Logistic regression analyses supported this association. Elevated neutrophil count, neutrophil-to-erythrocyte ratio, and neutrophil-to-HDL ratio were associated with an increased likelihood of having an IMT-CC >.9 mm {odds ratio (OR) 1.17 [95% confidence interval (CI) 1.04–1.35], OR 2.21 (95% CI 1.24–4.12), and OR 1.96 (95% CI 1.09–3.55), respectively}, after adjustment for all variables, which was corroborated by linear regression. Furthermore, a linear mixed-effect model analysis from a longitudinal analysis spanning 5 and 7 years revealed an increase in 1 unit of neutrophils/μl at these time points was associated with a mean increase of .004 (.002) mm in the IMT-CC (P = .031) after adjustment for all variables. Interestingly, in patients exhibiting regression in IMT-CC after 7 years of follow-up, those following a Mediterranean diet showed a significant decrease in neutrophil count after 5 and 7 years (both with P < .05), compared to baseline.

Conclusions

These findings suggest that neutrophils may represent a promising target for preventing atherosclerosis. A Mediterranean diet could serve as an effective dietary strategy to reduce neutrophil levels and potentially slow the progression of atherosclerosis, offering a new neutrophil-reducing therapy concept. Further research is essential to gain deeper insights into the role of neutrophils in the pathophysiology of atherosclerotic cardiovascular disease in humans.

Structured Graphical Abstract

Structured Graphical Abstract.

Increased neutrophils were associated with greater intima-media thickness of common carotid arteries (IMT-CC) both at baseline. Furthermore, the longitudinal analysis spanning 5 and 7 years revealed an increase in 1 unit of neutrophil value/μL at these time points is associated with a mean increase of .004 mm in the IMT-CC. Nonetheless, adherence to the Mediterranean Diet resulted in a reduction in neutrophil among patients who exhibited a decrease in IMT-CC after 7-year of follow-up, suggesting a novel neutrophil-reducing therapy strategy. IMT-CC, Intima-media thickness of common carotid arteries.

• Key Question: What is the clinical implication of neutrophils in atherosclerosis development over 5- and 7-year follow-up? Can dietary approaches, such as a Mediterranean diet, help reduce neutrophils and minimize this risk, suggesting a new therapeutic strategy?
• Key Findings: Increased neutrophils were associated with greater intima-media thickness of common carotid arteries (IMT-CC) both at baseline and after the longitudinal analysis. Nonetheless, adherence to the Mediterranean diet resulted in a reduction in neutrophils among patients who exhibited a decrease in IMT-CC after 7-year follow-up.
• Take Home Message: Neutrophils may serve as a therapeutic target for atherosclerosis prevention in humans. The Mediterranean diet, as a therapeutic strategy, could reduce neutrophils and slow atherosclerosis progression, supporting a novel neutrophil-reducing therapy plan.

See the editorial comment for this article ‘Connecting the dots: the Mediterranean diet, neutrophil count, and atherosclerosis’, by M. Guasch-Ferré, https://doi.org/10.1093/eurheartj/ehae799.

Introduction

Atherosclerotic cardiovascular disease (ASCVD) remains the leading cause of morbidity and mortality in developed countries, imposing a significant economic burden on healthcare systems and society at large.¹ Carotid intima-media thickness has been identified as a surrogate marker of global atherosclerosis. In fact, several studies have supported its association with both ASCVD incidence and prognosis. Accordingly, the global prevalence rates for increased intima-media thickness of both common carotid arteries (IMT-CC) have been estimated at 27.6% worldwide. In addition, the presence of carotid plaques (another marker of ASCVD) has been estimated to have a prevalence of 21.1% affecting ∼1 billion individuals.² Furthermore, the same report estimates an increase of ∼57% in these figures from 2000 to 2020, likely linked to lifestyle habits and unhealthy dietary patterns.³

In the context of ASCVD, inflammation has been widely recognized as a primary driver of the disease.⁴ Over the past few decades, an increased count of white blood cells has been well-recognized as a distinct risk factor for cardiovascular disease,⁵ and evidence regarding the significant role that neutrophils may play in cardiovascular disease has widely been discussed.⁶ In fact, a very recent randomized Mendelian study reported that high neutrophil count may be considered a causal risk factor for ASCVD.⁷ Supporting this, recent studies have demonstrated the presence of neutrophils in atherosclerotic lesions.⁸ Mechanistically, neutrophils participate in various stages of atherosclerosis development, as well as playing a central role in the destabilization of plaques.⁹ This is primarily achieved by recruiting and activating other immune cells within the atherosclerotic lesion, leading to the release of proinflammatory mediators and the induction of smooth muscle cell death, among others.^6,9,10 Although there is increasing evidence suggesting the mechanistic role of neutrophils of atherosclerosis, mainly observed in mice, their functions in human atherogenesis, the progression of atherosclerosis, and the destabilization of plaques remain poorly understood. In this context, several circulating biomarkers related to neutrophil count (as well as neutrophil-related ratios) have been proposed, to minimize the effect of concomitant conditions in the bloodstream that could affect the accuracy of neutrophil count, such as acute inflammation, haemoconcentration, and others.¹¹

On the other hand, previous findings from the CORDIOPREV study showed that dietary intervention influences IMT-CC.^12–14 However, the mechanisms underlying this effect remain unresolved. To the best of our knowledge, no study has examined the significance of neutrophil count on IMT-CC progression through dietary intervention. Therefore, in this study, we assessed whether neutrophil count and their related ratios could serve to evaluate an increase in IMT-CC at baseline, and whether long-term follow-up of two different healthy dietary models differentially interacts with neutrophils to determine IMT-CC progression over 7 years.

Methods

Study design and participants

This research project is a pre-specified secondary objective of the CORDIOPREV study registered on the clinical trials website (https://clinicaltrials.gov) under NCT00924937, which examined the effects of two different dietary approaches on the incidence of new cardiovascular events in individuals undergoing secondary prevention. In the CORDIOPREV study, a total of 1002 patients with coronary heart disease (CHD) were enrolled between 1 October 2009, and 28 February 2012, and were randomly assigned to follow either a low-fat diet or a Mediterranean diet for 7 years while continuing their conventional CHD treatment. The study's rationale, methodology, and baseline participant characteristics have been previously described in detail, and the main outcomes have been reported in The Lancet.¹⁵ The study protocol adhered to the principles of the Declaration of Helsinki and received approval from the Institutional Review Board of the Reina Sofia University Hospital (Córdoba, Spain). All participants provided written informed consent.

Dietary assessment of the participants

Within the CORDIOPREV study design, participants were randomly assigned to follow one of two dietary patterns: a low-fat diet [<10% of calories from saturated fat, 12%–14% from monounsaturated fatty acids (MUFAs), 6%–8% from polyunsaturated fatty acids (PUFAs), constituting <30% of calories from fat, 15% protein, and a minimum of 55% carbohydrates] or a Mediterranean diet (<10% of calories from saturated fat, 22% from MUFAs, 6% from PUFAs, with a minimum of 35% of calories from fat, 15% protein, and a maximum of 50% carbohydrates). The assessment of dietary intake has been previously described.¹⁶ All participants received comprehensive dietary guidance, including individual assessments (conducted at the study's start and every 6 months) and group educational sessions tailored specifically for each diet group (held every 3 months, with up to 20 participants in each session). Both diets were designed to include a variety of food options, encompassing vegetables, fruits, grains, potatoes, legumes, dairy products, meats, and fish. Participants’ habitual dietary habits were evaluated using a validated 137-item semi-quantitative food frequency questionnaire by registered dietitians at the study's start and at 6-month intervals throughout the follow-up period.¹⁷

Laboratory measurements

After an overnight fasting period, all participants provided blood samples. Serum was separated from the blood by centrifugation at 1500 x g for 15 min at 4°C. Fasting glucose, total cholesterol, triglycerides, and high-density lipoprotein (HDL) cholesterol levels were assessed in serum using the Dimension Autoanalyzer (Dade Behring Inc., Deerfield, IL, USA). Calculated low-density lipoprotein (LDL) cholesterol levels were determined using the Friedewald formula.¹⁸ Fasting insulin levels in plasma were measured using a chemiluminescent assay with the i-2000 Abbott Architect® analyzer. High-sensitivity C-reactive protein (hsCRP) concentrations in plasma were determined using a high-sensitivity enzyme-linked immunosorbent assay (BioCheck, Inc., Foster City, CA, USA). Glycosylated haemoglobin (HbA1c) levels were measured in fresh samples using the G8 analyzer (Tosoh Corporation, Shiba-Mianto-ku, Tokyo, Japan).

Flow cytometry with an automated analyzer was employed to assess the counts of various blood components, including erythrocytes, lymphocytes, monocytes, and neutrophils. Several ratios were calculated to analyse the relationships between neutrophils and other IMT-CC outcomes, including the following: the neutrophil-to-erythrocyte ratio (NER), which is determined by dividing the absolute counts of neutrophils by the absolute counts of erythrocytes. This marker serves to indicate systemic inflammatory states, alterations in the immune system, and various health conditions. The neutrophil-to-HDL ratio (NHR), calculated by dividing the absolute counts of neutrophils by the absolute counts of HDL. This biomarker is associated with chronic and systemic inflammation, as well as cardiovascular risk, as neutrophils signify inflammation while HDL is considered cardioprotective. The neutrophil-to-lymphocyte ratio (NLR), computed by dividing the absolute counts of neutrophils by the absolute counts of lymphocytes. Widely employed for assessing inflammation, cardiovascular disease, and immune response, a higher NLR indicates an increase inflammatory response.

Measurement of intima-media thickness of common carotid arteries

Bilateral ultrasound examinations of the carotid arteries were conducted using a high-resolution B-mode Doppler ultrasound device (Envisor C Ultrasound System, Phillips, USA) at baseline and after 5 and 7 years of dietary intervention. The ultrasound procedures followed the guidelines of the American Society of Echocardiography Carotid Intima-Media Thickness Task Force,¹⁹ as previously described by Jimenez-Torres et al.²⁰ Briefly, three measurements of the IMT-CC were taken for each common carotid artery (left and right) for each participant. The average of the IMT-CC measurements for each side was calculated independently, and the overall IMT-CC value was determined by averaging the IMT-CC measurements for both common carotid sides. All ultrasound images were analysed offline using dedicated analysis software (QLAB Advance Ultrasound Quantification Software, v5.0, Phillips, USA) to measure the IMT-CC. Out of 1002 patients, 939 completed the carotid ultrasound study at baseline, while the remaining 63 patients did not. At the 5-year follow-up, 809 patients completed the carotid ultrasound study (86% of the participants), while 130 patients did not complete the study at this point. Finally, at the 7-year follow-up, 731 patients completed the carotid ultrasound study (78% of the participants), while 78 patients did not. The main reasons for dropout were refusal to undergo carotid ultrasonography, death, or withdrawal for other reasons.

Statistical analysis

The results of the study are presented as means and standard deviations (SD) for continuous variables and as percentages for categorical variables. Student’s t-test or the Mann–Whitney test was used for comparisons, depending on the normality of the distribution (assessed by Shapiro–Wilk). For paired samples, the Student’s t-test or Wilcoxon test was used for comparisons in cases of non-normality. For comparisons between three groups, the ANOVA or Kruskal–Wallis tests were employed according to the normality test, followed by pairwise comparisons of P-values (two-sided), corrected for multiple comparisons. Logistic and linear models were conducted using data from the baseline. According to the European Society of Cardiology guidelines, logistic regression was conducted using an IMT-CC cut-off of ≥.70 mm to assess the relationship between neutrophil count and increased risk of cardiovascular diseases at baseline.^21–23 Additionally, an IMT-CC cut-off of ≥.90 mm was employed to evaluate the relationship between neutrophil count and organ damage at baseline.^24–26 All adjustment models were as follows: Model 1 was an unadjusted model. Model 2 (basic demographic variables) was adjusted for such as age, sex, and body mass index (BMI). Model 3 (cardiometabolic risk factors) was adjusted for age, sex, BMI, smoking, alcohol intake, LDL, the presence of Type 2 diabetes mellitus, and hypertension. Model 4 (dietary and lipid factors) was adjusted for age, sex, BMI, triglycerides, HDL, statins, and diet (low-fat vs. Mediterranean diet). Model 5 (pharmacological treatments) includes the Model 4, plus β-blockers, angiotensin-converting enzyme inhibitors (ACEIs), angiotensin receptor blockers (ARBs), diuretics, calcium antagonists, other lipid-lowering, antihypertensive drugs, and oral antidiabetics. Model 6 includes all variables. For the linear mixed-effect model, we investigated the relationship between the IMT-CC as the dependent variables (IMT-CC was measured at 0, 5, and 7 years) and their fixed effects, including neutrophils, diet, or time (variable) as independent variables over the follow-up period. Neutrophil count was also measured at 0, 5, and 7 years. All participants were undergone with low-fat or Mediterranean diet. In terms of time, the follow-up was at 0, 5, and 7 years. For random effects, an ‘inter-subject’ term was included in the model to evaluate the random variation in the intercept across different individuals. To assess multicollinearity in our models, we examined the variance inflation factor (VIF) using Car package for all variables included. No multicollinearity was found for neutrophil count and other related variables included in this study, with most VIF values close to 1 and some below 5. Progression of IMT-CC was calculated as positive change (IMT-CC at 7-years—IMT-CC at baseline >0) and regression of IMT-CC was calculated as unchanged or negative change (IMT-CC at 7-years—IMT-CC at baseline ≤0). All analyses and graphical representations were performed using R v.3.5.1 software, and statistical significance was set at P < .05 (two-sided).²⁷

Results

Baseline characteristics of the participants

The baseline characteristics of the participants are shown in Table 1. We divided the participants into three groups based on tertiles of neutrophil count. The mean neutrophil count for the lowest tertile (T1) were 2.90 (.45), for the middle tertile (T2) were 3.97 (.29), and for the highest tertile (T3) were 5.69 (1.74). The highest tertile group exhibited higher anthropometric variables, including weight (P < .05), waist circumference (P < .001), and BMI (P < .01) when compared to the lowest tertile group.

Table 1.

Baseline characteristics of the population divided by groups of tertiles baseline of neutrophils

Variables	T1 N = 346	T2 N = 322	T3 N = 334	P-value
Age (years)	58.8 (9.11)	60.1 (8.92)	59.9 (8.99)	.138
Sex, n (%)				.043
Male	272 (78.6%)	276 (85.7%)	278 (83.2%)
Female	74 (22.4%)	46 (15.3%)	54 (17.8%)
Systolic blood pressure (mm Hg)	139 (20.2)	139 (19.0)	138 (20.7)	.661
Diastolic blood pressure (mm Hg)	78.0 (11.2)b	78.2 (11.2)c	75.6 (10.1)	.003
Weight (kg)	83.4 (13.9)a,b	86.1 (14.6)	86.0 (14.6)	.010
Waist circumference (cm)	103 (11.0)a,b	105 (11.1)	107 (11.7)	<.001
Body mass index (kg/m²)	30.6 (4.37)a,b	31.2 (4.46)	31.6 (4.66)	.003
Total cholesterol (mg/dL)	161 (32.7)	160 (31.2)	156 (29.3)	.123
HDL cholesterol (mg/dL)	43.9 (10.3)a,b	42.1 (9.99)	40.5 (10.1)	<.001
LDL cholesterol (mg/dL)	91.0 (26.4)b	89.5 (24.2)	85.3 (26.3)	.014
Apolipoprotein A1 (mg/dL)	132 (21.3)b	130 (21.5)c	126 (21.8)	.001
Apolipoprotein B (mg/dL)	74.5 (19.4)	73.3 (17.5)	73.1 (17.8)	.822
Triglycerides (mg/dL)	128 (70.2)a,b	137 (64.9)	142 (73.1)	.002
Fasting glucose (mg/dL)	110 (35.2)b	113 (38.0)c	122 (52.9)	.006
Fasting insulin (µUI/mL)	10.2 (12.1)b	10.8 (9.90)c	12.1 (11.3)	.003
HOMA-IR	3.05 (2.82)b	2.91 (1.82)	3.38 (2.75)	.035
HbA1c (%)	6.44 (1.08)a,b	6.59 (1.17)c	6.99 (1.42)	<.001
Diabetes, n (yes, %)	148 (42.8%)a,b	170 (57.8%)c	220 (65.9%)	<.001
Hypertension, n (yes, %)	233 (67.3%)	226 (70.2%)	223 (66.8%)	.641
Smoking, n (%)				.001
Never	103 (29.8%)b	86 (26.7%)	75 (22.6%)
Former	224 (64.7%)	205 (63.7%)	207 (62.3%)
Current	19 (5.49%)	31 (9.63%)	50 (15.1%)
Alcohol intake, n (%)				.500
Non-drinkers	60 (17.5%)	49 (15.6%)	63 (19.2%)
8< g/day	144 (42.1%)	118 (37.5%)	129 (39.3%)
8–16 g/day	54 (15.8%)	67 (21.3%)	61 (18.6%)
>16 g/day	84 (24.6%)	81 (25.7%)	75 (22.9%)
Neutrophils (10³/µL)	2.90 (.45)a,b	3.97 (.29)c	5.69 (1.74)	<.001
Monocytes (10³/µL)	.40 (.19)a,b	.45 (.14)c	.59 (.32)	<.001
Lymphocytes (10³/µL)	1.86 (.57)a,b	2.00 (.59)c	2.21 (.66)	<.001
NLR	1.68 (.52)a,b	2.15 (.66)c	2.80 (1.16)	<.001
NER	.62 (.11)a,b	.84 (.10)c	1.21 (.41)	.001
NHR	.07 (.02)a,b	.10 (.02)c	.15 (.10)	<.001
hsCRP (mg/mL)	2.54 (3.39)a,b	2.82 (3.01)c	3.98 (4.31)	<.001
Baseline medication, n (yes, %)
Statins	291 (84.3%)	269 (83.5%)	295 (89.1%)	.086
β-Blockers	281 (84.1%)	259 (81.7%)	263 (80.7%)	.491
ACEIs	184 (55.1%)	161 (50.8%)	170 (52.1%)	.530
ARBs	86 (25.7%)	96 (30.3%)	111 (34.0%)	.066
Diuretics	121 (36.2%)	133 (42.0%)	144 (44.2%)	.100
Calcium antagonist	51 (15.3%)	64 (20.2%)	67 (20.6%)	.150
Other lipid-lowering drugs	303 (90.7%)	290 (91.5%)	302 (92.6%)	.670
Antihypertensive drugs	308 (89.0%)	287 (89.1%)	304 (91.8%)	.389
Oral antidiabetics	88 (25.4%)a,b	108 (33.5%)c	152 (45.8%)	<.001

Data are represented as mean (SD) or n (%). Groups were divided according to the baseline tertiles of absolute neutrophil counts. A Shapiro–Wilks test was conducted to determine whether the data followed a normal distribution. ANOVA test or the Kruskal–Wallis test was used depending on the normality of the variables. Pairwise comparisons were then performed with P-values adjusted for multiple testing. For variables expressed as percentages, the χ² test was applied. Statistical significance is indicated as follows: P-values <.05 have been bolded.

Patients with Type 2 diabetes mellitus include patients who either self-reported a prior diagnosis or met the American Diabetes Association diagnostic criteria, which include an HbA1c level of 6.5% or higher, a fasting blood glucose level of 126 mg/dL or above, or a 2-h blood glucose level of 200 mg/dL or higher following a 75 g oral glucose tolerance test conducted at baseline.

ACEIs, angiotensin-converting enzyme inhibitors; ARBs, angiotensin II receptor blockers; g, grams; HbA1c, haemoglobin glycosylated; HDL, high-density lipoprotein; HOMA-IR, homeostasis model assessment-insulin resistance; hsCRP, high-sensitive C-reactive protein; LDL, light density lipoprotein; NER, neutrophil-to-erythrocyte ratio; NHR, neutrophil-to-HDL ratio; NLR, neutrophil-to-lymphocyte ratio; T1, tertile 1, lowest tertile; T2, tertile 2, middle tertile; T3, tertile 3, highest tertile.

^aAdjusted P-value: T1 vs. T2.

^bAdjusted P-value: T1 vs. T3.

^cAdjusted P-value: T2 vs. T3.

Additionally, the highest tertile group showed decreased lipid variables, such as HDL-c (P < .001), LDL-c (P < .05), APOA1 (P < .01), but increased triglyceride levels (P < .01), when compared to the lowest tertile group. Furthermore, the highest tertile group showed differences in markers of glucose metabolism and insulin sensitivity, such as fasting insulin and glucose levels (both with P < .01), homeostasis model assessment-insulin resistance (HOMA-IR) (P < .05), HbA1c (P < .001), as well as inflammatory markers, such as hsCRP (P < .001), when compared to the lowest tertile group. Finally, all neutrophil-related ratios were higher in the highest tertile group compared to both middle and lowest tertile groups (all with P < .001). Multiple comparisons between all tertiles are summarized in Supplementary data online, Table S1.

Neutrophil count and the risk of increased intima-media thickness of both common carotid arteries at baseline

To investigate the relationship between neutrophils and IMT-CC at baseline, we compared participants across tertiles of neutrophil count. We found that the highest tertile showed significantly increased IMT-CC compared to the lowest tertile (P = .003), although there was only a trend when comparing the middle tertile to the lowest (P = .079) and to the highest tertile (P = .075) (Figure 1A). Additionally, the highest tertile also showed a significantly higher number of plaques compared to the lowest tertile (P = .011) (Figure 1B). However, no significant differences were observed between the lowest and middle tertiles, nor between the middle and highest tertiles in terms of plaque count.

Figure 1.

Association between absolute neutrophil count and common carotid artery intima-media thickness. Asterisks indicate significant differences between the groups according to the Kruskal–Wallis test and further pairwise comparisons of P-values, corrected for multiple comparisons (*P < .05, **P < .01, ***P < .001). (A) Comparison of tertiles of baseline absolute neutrophil count and IMT-CC values (N = 939). After Kruskal–Wallis test, the P-value was .008, and the multiple comparison P-values were: T1 vs. T2 = .079; T1 vs. T3 = .003; and T2 vs. T3 = .075 (N = 939). (B) Comparison of tertiles of baseline absolute neutrophil count and total number of plaque values. After Kruskal–Wallis test, the P-value was .028, and multiple comparison P-values were: T1 vs. T2 = .120; T1 vs. T3 = .011 and T2 vs. T3 = .112 (N = 710). IMT-CC, intima-media thickness common carotid; T1, lowest tertile; T2, middle tertile; T3, highest tertile

To further explore the association between baseline neutrophil count and IMT-CC, a logistic regression was performed, using two pre-specified IMT-CC cut-offs, as outlined in the Materials and Methods Section. When applying the more severe cut-off of .9 mm, our analysis showed that neutrophil count was associated with an increased likelihood of elevated IMT-CC in both the unadjusted and fully adjusted models {odds ratio (OR) 1.24 [95% confidence interval (CI) 1.10–1.39] and OR 1.17 (95% CI 1.04–1.35), respectively}. Similar results were observed for NER [OR 2.70 (95% CI 1.60–4.63) and OR 2.21 (95% CI 1.24–4.12), respectively] and NHR [OR 2.42 (95% CI 1.64–3.59) and OR 1.96 (95% CI 1.09–3.55), respectively] (Table 2). Results for NLR and monocytes are summarized in Supplementary data online, Table S2.

Table 2.

Logistic regression analysis of the risk of intima-media thickness of common carotid arteries ≥.9 mm based on neutrophil variables

Variables	Models	OR (95% CI)	P-value
		IMT-CC ≥ .9 mm
Neutrophils	Model 1 (unadjusted)	1.24 (1.10–1.39)	<.001
	Model 2	1.23 (1.09–1.39)	<.001
	Model 3	1.19 (1.05–1.36)	.009
	Model 4	1.19 (1.06–1.35)	.005
	Model 5	1.20 (1.07–1.36)	.004
	Model 6	1.17 (1.04–1.35)	.018
NER	Model 1 (unadjusted)	2.70 (1.60–4.63)	<.001
	Model 2	2.67 (1.57–4.62)	<.001
	Model 3	2.35 (1.32–4.31)	.005
	Model 4	2.36 (1.39–4.14)	.002
	Model 5	2.31 (1.35–4.07)	.003
	Model 6	2.21 (1.24–4.12)	.011
NHRa	Model 1 (unadjusted)	2.42 (1.64–3.59)	<.001
	Model 2	2.32 (1.56–3.49)	<.001
	Model 3	2.01 (1.28–3.21)	.003
	Model 4	2.15 (1.27–3.70)	.005
	Model 5	2.05 (1.36–3.12)	<.001
	Model 6	1.96 (1.09–3.55)	.026

Logistic regression analysis using IMT-CC as categoric variable using IMT-CC ≥ .9 mm as cut-off value. Model 1 was an unadjusted model that includes only the independent variable. Model 2 was a basic adjusted model for basic demographic variables such as age, sex, and BMI. Model 3 was a model adjusted for cardiometabolic risk factors, such as age, sex, BMI, smoking, alcohol intake, LDL, the presence of type 2 diabetes mellitus and hypertension. Model 4 was a model adjusted for dietary and lipid factors, such as age, sex, BMI, triglycerides, HDL, statin, and diet (low-fat vs. Mediterranean diet). Model 5 was a model adjusted for pharmacological treatments, including the Model 4, and β-blockers, ACEIs, ARBs, diuretics, calcium antagonist, other lipid-lowering and antihypertensive drugs, and oral antidiabetics. Model 6 was a fully adjusted model, which include all variables used in the previous models, such as demographic, cardiometabolic, dietary and lipid factors as well as pharmacological factors. P-values <.05 have been bolded and indicate a statistically significant association between the dependent and independent variables.

ACEIs, angiotensin-converting enzyme inhibitors; ARBs, angiotensin II receptor blockers; BMI, body mass index; CI, confidence interval; IMT-CC, intima-media thickness of both common carotid arteries; HDL, high-density lipoprotein; OR, odds ration; LDL, light density lipoprotein; NER, neutrophil-to-erythrocyte ratio; NHR, neutrophil-to-HDL ratio.

^aSymbol means logarithm.

When using the .7 mm cut-off in the unadjusted model, we observed that neutrophil count was similarly associated with an increased likelihood of elevated IMT-CC [OR 1.14 (95% CI 1.03–1.26)], with comparable findings for NER, NHR and NLR. However, significance for these associations was lost in the adjusted models, although the trends remained in the same directions (see Supplementary data online, Table S3). Through linear regression analysis, we further corroborated a positive association between neutrophil count and IMT-CC at baseline in both the unadjusted (β (SD) = .015 (.003); P < .001) and the fully adjusted model (β (SD) = .010 (.003); P < .01). Additionally, similar patterns were observed for NER and NHR, though not for NLR and monocytes (see Supplementary data online, Table S4).

Increase in neutrophil count are associated with increase in intima-media thickness of both common carotid arteries after a follow-up of 5 and 7 years

To examine the evolution of neutrophils and IMT-CC over the follow-up period, we conducted a linear mixed-effects model. This model assessed the longitudinal association between IMT-CC and neutrophil count, accounting for the effects of diet and repeated measurements at different time points (0, 5, and 7 years). The model includes both fixed effects [neutrophils, diet, and time (variable)] and random effects (variations across individuals). All the analyses were adjusting for all variables used in this study.

We found that neutrophil count had a significant positive correlation with the IMT-CC [β (SE) = .004 (.002), P = .031], indicating that an increase in 1 unit of neutrophil value/μl corresponded to an average increase of .004 mm in IMT-CC. Conversely, allocation to a Mediterranean diet, however, was significantly associated with a reduction in IMT-CC [β (SE) = −.028 (.009), P = .001], reflecting an estimated decrease of .029 mm compared to a low-fat diet. Time (intervention time, as variable) also had a notable effect [β (SE) = −.004 (.001), P < .001] on the IMT-CC, showcasing an average reduction of .004 mm/year, likely due to the dietary intervention.

Regarding interactions, the neutrophil-diet interactions were not significant (P = .240). However, interaction between neutrophil count with time (variable) was significant, indicating that a decrease in 1 unit of neutrophil value/μl interacting with time (variable) was associated with a mean decrease of .002 mm in IMT-CC during the follow-up period [β (SE) = −.002 (.001), P = .002] (Table 3).

Table 3.

Linear mixed model analysis using intima-media thickness of both common carotid arteries and neutrophil counts

Variables	Estimate	Standard error	P-value
Individual model
Neutrophils	.004	.002	.031
Diet	−.028	.009	.001
Time	−.004	.001	<.001
Interaction model
Neutrophils vs. Diet	−.006	.005	.240
Neutrophils vs. Time	−.002	.001	.002

Linear mixed-effects model analysis was designed to investigate the relationship between the IMT-CC as the dependent variables (IMT-CC was measured at 0, 5, and 7 years) and fixed effects, including Neutrophils, Diet, or Time (variable) as independent variables over the follow-up time. Neutrophils were also measured at 0, 5, and 7 years. All participants were undergone with low-fat or Mediterranean diet. In terms of Time, the follow-up was at 0, 5, and 7 years, as previously mentioned. For random effect (1 | patients), we used this syntax to evaluate the random variation in the intercept across different individuals. This allows for the possibility that the baseline IMT-CC may vary between individuals, even after accounting for the fixed effects. For the individual model, Neutrophils, Diet, or Time (variable) were adjusted for all variables [including age, sex, BMI, smoking, alcohol intake, LDL, the presence of Type 2 diabetes mellitus and hypertension, triglycerides, HDL, statin, and diet (low-fat vs. Mediterranean diet), β-blockers, ACEIs, ARBs, diuretics, calcium antagonist, other lipid-lowering and antihypertensive drugs, and oral antidiabetics]. For the interaction model, Neutrophils, Diet, and Time were included in the interaction model, as well as were adjusted for all variables. P-value <.05 have been bolded.

IMT-CC, intima-media thickness common carotid.

Mediterranean diet is associated with a decrease in neutrophil count in those patients who showed a decrease in intima-media thickness after a 7-year follow-up

To understand how neutrophils interact with time (variable) and diet, and their association with IMT-CC after a 7-year follow-up, we categorized patients based on whether their IMT-CC progressed (progression group) or regressed (regression group) during this period. Among patients where IMT-CC progressed, neutrophil count decreased significantly after a 5-year follow-up (P = .001); however, this change was not significant when comparing baseline to the 7-year mark (P = .126). In contrast, in the subgroup with IMT-CC regression, neutrophil count decreased significantly at both after 5 and 7 years (P = .005 and P = .019, respectively), when compared to baseline (Figure 2).

Figure 2.

Association between absolute neutrophil count and intima-media thickness of common carotid arteries. Asterisks indicate significant differences between the groups according to the Wilcoxon test for paired samples (*P < .05, **P < .01, ***P < .001). (A) Comparison of neutrophil count between 0 and 5 years and between 0 and 7 years in those patients who experienced a progression (increase in intima-media thickness of common carotid arteries; N = 304) and regression (decrease in intima-media thickness of common carotid arteries; N = 409) of intima-media thickness of common carotid arteries between 0 and 7 years. (B) Comparison of neutrophil count between 0 and 5 years and between 0 and 7 years in those patients who experienced regression (decrease in intima-media thickness of common carotid arteries) of intima-media thickness of common carotid arteries between 0 and 7 years and followed either low-fat diet (N = 170) or a Mediterranean diet (N = 239). IMT-CC, intima-media thickness of common carotid arteries

Next, we further focused on and stratified the regression group on their adherence to either the Mediterranean or low-fat diet to examine the influence of the diet on this association. In the low-fat diet group, neutrophil count remained stable over the years, as measured at both the 5- and 7-year follow-up points (P = .112 and P = .382, respectively). However, in the Mediterranean diet group, neutrophil count decreased significantly at both time points (5 and 7 years) compared to baseline (P = .024 and P = .021).

Discussion

This study presents new findings on the role neutrophils in the development of atherosclerosis. We found a strong positive association between neutrophil count and IMT-CC, consistent across multiple statistical analyses, including linear, logistic, and linear mixed model. Over a 5- to 7-year follow-up, we observed a parallel increase in both neutrophil count and IMT-CC, suggesting that neutrophils could potentially serve as a valuable biomarker for assessing increased IMT-CC. In contrast, patients who experienced regression in IMT-CC after 7 years also experimented a reduction in neutrophil count at both 5 and 7 years, particularly those following a Mediterranean diet (Structured Graphical Abstract). These findings suggest that targeting neutrophils could be a promising strategy for preventing atherosclerosis. A Mediterranean diet may be especially effective in lowering neutrophils and potentially slowing the progression of atherosclerosis, presenting a novel concept for neutrophil-reducing therapy.

Several studies have identified a positive association between neutrophil count and IMT-CC,^28,29 as well as with other neutrophil-related ratios, such as NLR.^30–32 The NLR, in particular, has been also prospectively linked to cardiovascular mortality, cardiovascular diseases, and heart failure in the Jackson Heart Study.³³ In our study, we evaluated the association between neutrophils and IMT-CC, along with other related ratios, such as NLR, NER, and NHR in a large and long-term secondary prevention study. For this purpose, we used both .7 and .9 mm cut-offs to assess the relationship between neutrophils and IMT-CC. After adjusting for all covariates, no association was found between neutrophil count and IMT-CC at the .7 mm cut-off, suggesting that neutrophils may not play a clear role in the atherosclerosis initiation. However, at the .9 mm cut-off, we found that neutrophils were strongly associated with increased IMT-CC, which may suggest that increased neutrophils could serve as a risk factor for atherosclerosis progression, and may have clinical utility after further validation. Nonetheless, it is important to note that the application of these cut-offs still remains a matter of debate. They are somewhat arbitrary and show relative variability across populations with different demographic and health profiles, leading to inconsistencies in identifying at-risk individuals, and limiting the generalizability of our findings.³⁴ Future research should explore alternative thresholds, employ continuous analysis, and include studies across diverse populations.³⁵

Nevertheless, a recent observational and genetic study determined that high blood neutrophil count is a causal risk factor for ASCVD,⁷ underscoring the need to unravel potential mechanistic mechanisms behind this relationship. In this sense, recent studies in mice models also support the role of neutrophils in the progression of atherosclerosis. Neutrophils have been linked to the initiation of atherosclerosis,^8,36 through mechanisms such as monocyte recruitment,³⁷ dysregulation and activation of endothelial cells,³⁸ and the activation of macrophages via NETosis.³⁹ Additionally, neutrophils contribute to atherosclerotic plaque instability, promote death of smooth muscle endothelial cell, and cause endothelial damage within atherosclerotic plaques.^9,10,40 Therefore, these findings indicate a pivotal role of neutrophils in the progression of atherosclerosis and the onset of cardiovascular events. However, despite these findings, further validation in humans is limited but crucial, particularly concerning the role of neutrophils in both early and late-stage atherosclerosis.

Our study is the first to evaluate the impact of neutrophils on IMT-CC progression in the context of long-term dietary interventions. We found that that increase in neutrophil count were associated with IMT-CC progression over the follow-up period. However, among patients who experimented IMT-CC regression after 7 years, those following a Mediterranean diet showed a significant decrease in neutrophil count over 5 and 7 years of follow-up. In alignment with these findings, the Moli-sani study was the first to demonstrate that adherence to a Mediterranean diet was associated with reduced leukocyte counts, partially due to the high dietary fibre and antioxidant content.⁴¹ Similarly, Hernaez et al. reported within the PREDIMED study that the Mediterranean diet decreased the incidence of white blood cell count-related alterations in individuals at high cardiovascular risk.⁴² Particularly, the Framingham Heart Study identified a correlation between higher diet quality and reduced neutrophil count.⁴³ This effect was further reinforced by a study in which participants following a Mediterranean diet supplemented with omega-3 (as an anti-inflammatory treatment), exhibited a decrease in the NLR ratio.⁴⁴ These data suggest that Mediterranean diet may regulate granulopoiesis in humans, although the underlying mechanisms are still under research.⁴⁵ Potential mechanisms may include a combination of actions, such as the regulation of hyperglycaemia and hypercholesterolaemia, the reduction of cytokine production,⁶ and the modulation of neutrophil phenotype.^46,47 In summary, while the positive effects of the Mediterranean diet might be partially attributed to the regulation of granulopoiesis and neutrophil function, this study represents the first instance to purpose the Mediterranean diet as a potential neutrophil-reducing therapy for preventing related diseases in humans.

Several limitations should be considered concerning this study. Firstly, this research is based on a prolonged and carefully regulated dietary intervention, which may not precisely mirror real-life scenarios where individuals make dietary choices without constraints. Secondly, although this is a pre-specified secondary objective of the CORDIOPREV study, this was not the primary objective, hence requiring cautious interpretation of our findings. Further studies emphasizing IMT-CC and the predictive value of neutrophils as the primary outcome are essential for validation. Other limitation of our study is the impact of participant dropouts and competing risk events on our findings. The dropout rates were of 10% and 22%, at the 5- and 7-year follow-up, respectively. Although these percentages of dropout were relatively low, it is important to note, that could potentially affect the interpretation of our study. The loss of participants may lead to a reduction in statistical power, introduce bias and limit the generalizability of the results. Therefore, the potential impact of these dropouts should be considered for interpretation. However, we ensured a meticulous adjustment for main factors including demographic, cardiovascular, dietary and lipid factors, as well as pharmacological treatment to control all clinical outcomes and cofounding variables. An important observation in our study was the change in statistical significance when adjusting for potential confounders in our analyses. This finding underscores the significance of considering and accounting for confounding variables in interpreting the results. Lastly, there is a necessity for additional molecular and mechanistic investigations into the correlation between neutrophils and IMT-CC and their contribution to the initiation of atherosclerosis, particularly in the context of dietary intervention using the Mediterranean diet as well as their components. These additional studies would provide a more comprehensive understanding, potentially shedding light on the observations made in this research.

Conclusion

Our study reveals a significant link between neutrophil count and the progression of atherosclerosis. We found a strong positive association between neutrophil count and IMT-CC, consistently observed across multiple statistical analyses. Over a 5-to-7-year follow-up period, we observed a parallel rise in neutrophil count and IMT-CC. Conversely, patients who experienced a regression in IMT-CC after 7 years also showed a reduction in neutrophil count at both 5- and 7-year marks, especially those following a Mediterranean diet. These findings suggest that neutrophils could be a promising strategy for preventing atherosclerosis. The Mediterranean diet, in particular, may help reduce neutrophil levels and potentially slow the progression of atherosclerosis, introducing the concept of a neutrophil-reducing therapy. Further research is imperative to better understand the role of Mediterranean diet in modulating neutrophils and its impact on the pathophysiology of ASCVD.

Supplementary data

Supplementary data are available at European Heart Journal online.

Declarations

Disclosure of Interest

All authors declare no disclosure of interest for this contribution.

Data Availability

The data underlying this article cannot be shared publicly for the privacy of individuals that participated in the study. The data contain sensitive patient information. The data will be shared on reasonable request to the corresponding author.

Funding

The CORDIOPREV study is supported by the Fundacion Patrimonio Comunal Olivarero, by Junta de Andalucia (Consejeria de Salud, Consejeria de Agricultura y Pesca, Consejeria de Innovacion, Ciencia y Empresa CVI-7450 to J L-M), Diputaciones de Jaen y Cordoba, Centro de Excelencia en Investigacion sobre Aceite de Oliva y Salud and Ministerio de Medio Ambiente, Medio Rural y Marino, Gobierno de España. The CORDIOPREV study has also received research grants AGL2012/39615, AGL2015-67896-P, PID2019-104362RB-I00 and PID2023-1468910B-I00 funded by MCIN/AEI/1013039/501100011033 to J.L.M. and FIS PI10/01041 to P.P.M., FIS PI13/00023 to J.D.L. integrated into the framework of the National Plan for Scientific Research, Technological Development and Innovation 2013–2016, co-financed by the Instituto de Salud Carlos III (ISCIII) of Spain, the Directorate General for Assessment and Promotion of Research and the EU’s European Regional Development Fund (FEDER). HB is supported by a ‘Sara Borrell’ postdoctoral contract (CD22/00053) from the Instituto de Salud Carlos III - Madrid (Spain), ‘Financiado por la Unión Europea – NextGenerationEU’ y mediante el Plan de Recuperación, Transformación y Resiliencia and by M-AES mobility program with reference MV23/00060. O.S. receives support from the Deutsche Forschungsgemeinschaft (TRR332 projects A2 & Z2, SFB1123 project A6, SFB1009 project B13, KFO342 project B1), Novo Nordisk, the IZKF, and the IMF Münster as well as the EU Horizon Europe (‘Praetorian Network’) program. This study was supported by ‘Centro de Investigacion Biomédica en Red Fisiopatología de la Obesidad y Nutricion’, which is an initiative of the ‘Instituto de Salud Carlos III’ (ISCIII) of Spain, financed by the European Regional Development Fund, ‘A way to make Europe’/‘Investing in your future’.

Ethical Approval

The study protocol adhered to the principles of the Declaration of Helsinki and received approval from the Institutional Review Board of the Reina Sofia University Hospital (Córdoba, Spain). All participants provided written informed consent.

Pre-registered Clinical Trial Number

This research project is a pre-specified secondary objective of the CORDIOPREV study registered on the clinical trials website (https://clinicaltrials.gov) under NCT00924937.