Case selection

A total of 390 carotid specimens from the Interinstitutional Carotid Tissue Bank (ICTB) [18] were studied; carotid samples were collected from 265 symptomatic patients (major stroke or transient ischemic attack; 67.9%) and 125 asymptomatic patients (32.1%) underwent to surgical carotid endarterectomy (CEA) at the University of Rome Tor Vergata (Italy).

Among symptomatic patients, only those diagnosed with thrombo-embolism due to carotid atherosclerosis were included, whereas patients referred to any other cause of thrombo-embolism assessed by clinical examination and imaging were excluded. Specifically, cerebral CT scan study and angiographic examination of extra and intracranial carotid arteries and their branches were performed as routine pre-operative evaluation in all patients. Exclusion factors from the study were: (1) a possible cardiac source of embolization (rhythm disorders, recent myocardial infarction, stenosis, prolapse or calcification of mitral valve, left ventricular thrombus, mechanical cardiac valves, atrial myxoma, endocarditis, dilated cardiomyopathy, a patent foramen ovale), (2) stenosis greater than 50% of Willis circle.

All asymptomatic patients showed a carotid stenosis of at least 60%. For each patient, one carotid sample was collected. Informed consent was obtained from all subjects enrolled in this study. Study protocol adhered to ethical guidelines of the 2000 Declaration of Helsinki and was granted an a priori approval by the IRBs of our Institution.

Risk factors

Patients with a BMI ≥ 30.0 kg/m² were considered as obese, as defined by the American Heart Association (AHA) criteria [13]. Additional cardiovascular risk factors were defined as follows: (a) hypertension: systolic BP ≥ 140 mmHg and/or a diastolic BP ≥ 90 mmHg corresponding to stage 1 of JNC7 [21] and stage 2 of 2017 ACC/AHA guidelines [22]. Any patient under previous treatment with antihypertensive drugs was considered as being included under the diagnosis of hypertension; (b) diabetes mellitus: fasting blood glucose > 126 mg/dL and/or oral glucose-lowering treatment and/or insulin therapy; (c) smoke: patients with tobacco dependence as well as those who had stopped smoking for < 5 years; (d) metabolic syndrome, as defined in the AHA scientific statement [1]. In order to evaluate the dyslipidemia [23] low-density lipoprotein cholesterol (LDL-C) was calculated by the Friedewald equation [24]: LDL-C = cholesterol – [HDL-C + (triglycerides/5)]. A value of LDL-C of > 100 mg/dL was utilized as cut-off between high and low levels. Treatment with statins was also used as a criterion for inclusion in this category.

Histology

Sampling and histological methods have been previously reported [18, 25, 26]. Intraoperatively, carotid plaques were removed in block, immediately fixed in 10% buffered formalin, transversely cut every 5 mm, paraffin-embedded, and stained with hematoxylin–eosin and Movat stain. The entire plaque was evaluated by histology.

Plaques were classified into two categories according to the modified AHA atherosclerosis classification [27]: stable plaques and unstable plaques. Unstable plaques included those with cap rupture associated with acute thrombosis or with organized thrombi as well as vulnerable plaques or thin-cap fibroatheromas (TCFAs), the latter being characterized by the presence of a fibrous cap < 165 µm-thick and a marked infiltration by CD68 positive macrophages (at least 25/hpf), with no plaque rupture. Stable plaques included fibroatheromatous plaques with thick fibrous caps (> 165 µm) associated with the presence of calcification and a variable necrotic core [28], as well as healed plaques, defined as those showing multilayers of fibrous tissue.

Statistical analysis

Data were analyzed using SPSS version 16.0 (SPSS Inc, Chicago, Ill) software. Continuous variables were expressed as the mean ± SD. The normal distribution of the data was assessed by the Shapiro–Wilk test. Continuous variables were compared using the independent Student t-test or the Wilcoxon rank sum test. Categorical data were analysed using the Chi square test or the Fisher exact test.

Multivariate analysis using stepwise logistic regression (using the “enter” method for variable selection) was utilized to identify independent risk factors which significantly correlate with the presence of cerebrovascular symptoms (symptomatics vs. asymptomatics) or with an unstable carotid plaque. Multivariate analysis was performed in two models. One model was adjusted for the following parameters: age, gender and all the risk factors associated to metabolic syndrome (hypertension, diabetes, smoking habit, dyslipidemia (LDL-C), obesity. A second model was adjusted for age, gender, smoking habit, high cholesterol levels and metabolic syndrome, with or without obesity.

For each of the independent variables in the model, logistic regression was also used to estimate the coefficient (B), the estimated odds ratio (OR) [exp(B)] and the 95% confidence interval for exp(B) The statistical analyses were performed considering (a) all patients, (b) the male and female subgroups, (c) two cohorts of male patients divided according to the value of the median of age, i.e. 70 years.

A 2-tailed p value < 0.05 was considered statistically significant.

Baseline data

All patients had at least one risk factor. Hypertension (in 69.5% of cases) and dyslipidemia (in 65.9% of patients) were the risk factors most frequently observed. In particular, 43 patients (11.0%) showed a BMI > 30 kg/m² (obesity). Metabolic syndrome was observed in 123 cases (31.5%) and in 39 of these obesity constituted one of its components. Treatment with statins and diuretics was administered to 37.7 and 57.2% of patients, respectively. All patients in the post-operative and follow-up periods were treated with aspirin (100 md/die).

The histological examination showed in 253 patients (64.9%) the presence of unstable plaque (with acute thrombosis and/or TCFA), whereas in 137 (35.1%) the presence of the stable one (Fig. 1). Unstable plaques were significantly associated with the presence of symptoms (p = 0.001) (Table 1). The only unstable plaques observed in the asymptomatic group were represented by TCFAs.

Correlation among risk factors, symptoms and type of plaque

Multivariate analysis confirmed these results and showed an OR (increased risk of having an unstable plaque) of 2.58 (95% CI 1.08–6.19) for obesity. When the presence of metabolic syndrome was considered in the multivariate analysis, the latter had a highly significant independent effect (p = 0.001) with an OR of 2.55 (95% CI 1.61–4.90). If we considered patients with metabolic syndrome and obesity, the value of OR increased to 3.97 (95% CI 1.81–6.22) compared to the value of 1.48 (95% CI 0.86–2.31) when obesity was not included in the metabolic syndrome (Table 2).

Correlation among risk factors, gender and type of plaque

For male patients the risk factors significantly associated with the presence of unstable carotid plaques were: age (p = 0.03), hypertension (p = 0.001) and obesity (p = 0.01). However, obesity showed the higher OR (= 5.06, 95% CI 1.42–18.11) as compared to those of other risk factors.

In the female patients obesity was not significantly associated to the presence of unstable plaques (p = 0.92). In this subgroup of patients only hypertension was predictive of carotid plaque instability (p = 0.005).

Obesity and incidence of cardio and cerebrovascular diseases

In the last decade, several studies have reported an unexpected and seemingly paradoxical inverse correlation between BMI and incidence of cardiovascular disease. Extensive clinical data from patients with congestive heart failure and ischemic heart disease seem to support the hypothesis that overweight and obese individuals benefit from a more favorable short- and long-term prognosis [2, 7–9]. These results have been corroborated by autopsy-related and histologic data revealing a lower incidence of coronary atherosclerosis in obese decedents [42, 43]. According to Brodsky et al. [10] morbid obesity seems to exert a protective effect also against developing severe aortic atherosclerosis, even if the pathogenetic mechanisms for such observation are yet to be explained.

On the contrary, other clinical studies failed to show the protective effects of obesity. The LEADER trial demonstrated a prevalence and cardiometabolic impact of obesity in cardiovascular high-risk patients with type 2 diabetes mellitus [44]. In addition, the Partners registry has demonstrated that compared to normal BMI there was an increased burden of CAD for BMI > 25 kg/m² [45]. According to these data, the results of our study do not seem to confirm such paradoxical effect of obesity on the cerebral circulation, especially in the cohort of male patients with an age < 70 years; additionally, a neutral influence seems to exist among female patients (Table 4).

The discrepancy between our results and those obtained with a similar methodology in other vascular districts could be due to the fact that our study did not include the general population, but was rather limited to patients undergoing CEA due to symptomatic disease or presence of a carotid stenosis > 60%. Accordingly, our study did not evaluate the absolute incidence of atherosclerotic disease, but rather was focused on the unstable atherosclerotic plaque since it represents the target for prevention studies.

Obesity and aging

Another significant clinical result from our study is the observation that risk of carotid plaque destabilization in obese patients is linked to age. Indeed, OR was highest in male patients aged < 70 years. It was not possible to evaluate whether the risk increased in even younger patients as in our cases only 38 individuals were younger than 60 years. Despite this limitation, our results confirm data from other clinical studies reporting an inverse age-dependent association of BMI with all-cause mortality risk, including stroke mortality [14, 46]. In this context, the Global BMI Mortality Collaboration study recently reported a detailed analysis about the correlation between BMI and all-cause mortality, including atherosclerotic disease, showing that BMI was greater in younger than older people and in men than women [47]. Furthermore, prevention of hypertension, obesity, and diabetes in patients aged 44–56 years may substantially prolong heart failure-free survival and decrease heart failure-related morbidity [48, 49].

Study limitations

Another possible limitation of our study is the criterion used to define obesity, since BMI, waist-to-hip ratio, or waist circumference have been employed to evaluate the presence of obesity in different studies. Data in the literature in this regard are conflicting. According to some authors, abdominal body fat might represent a stronger predictor of stroke risk than BMI [50]. Other studies reported that only BMI was significantly associated with stroke in male patients, whereas for female individuals the ideal criterion to define obesity appeared to be waist-to-hip ratio [51]. In our study the AHA definition of obesity was used [13], as we believe that BMI appears to be the most frequent anthropometric index employed to measure the obesity in the available literature [52, 53].

Moreover, mixing the whole patient population (symptomatic and asymptomatic) may have determined a bias in the results, since patients who already had a stroke also presented a different risk factor profile overall. It was not possible to perform separately the multivariate statistical analysis in the group of symptomatic and asymptomatic patients as the latter was only constituted of 125 cases with only 6 obese, 4 with TCFA and 2 with stable plaques. To overcome this limitation we evaluated the possible correlation of obesity with the instability of the carotid plaque which represents the true target for stroke prevention.