Accuracy of Different Metabolic Syndrome Denitions for Detecting Increased Carotid Intima-media Thickness in Peruvian Type 2 Diabetes Mellitus Subjects without Cardiovascular Disease.

Backrgound: To estimate the accuracy of metabolic syndrome denitions for detecting Carotid Intima-Media thickness (CIMT) in Peruvian type 2 diabetes mellitus (T2DM) subjects without cardiovascular events. Methods: We performed a cross-sectional evaluation of T2DM subjects from the endocrinology service of two reference health centers without stroke or coronary disease. . Bilateral carotid intima-media thickness was measure by B mode ultrasound one only operator. We performed four denitions of Metabolic Syndrome; 1. Cholesterol Education Program Adult Treatment Panel III (ATP III), 2. Harmonized criteria; 3. Gurka's Metabolic syndrome severity score (MSSS) formula. 4. Non-glucose modied MSSS. We calculated the area under the receiver operator curve (AUROC) for detecting increased CIMT ( ≥ 0.86 mm) between metabolic syndrome denitions. Results: We included 184 subjects with T2DM, 29% were men with a mean age of 61.5 ± 10.5 years old. Median diabetes time was 10.8 years (IQR 4.8 to 19.4), and 26.8% achieved HbA1c goals. Non-glucose modied MSSS was the only denition signicantly correlated with elevated CIMT (r = 0.19; p<0.01), and it showed the best accuracy for predicting elevated CIMT (AUC: 0.61, CI95%: 0.52-0.70). Adjusting for age, sex, and HbA1c, each point increase in the non-glucose MSSS Z score, the risk of altered CIMT increases by 59% RP 1.59 (CI 95% 1.09 - 2.35; p=0.017). Conclusions: Non-glucose modied MSSS had a weak accuracy for elevated CIMT and was the best compared to ATP III, harmonized, and original MSSS. Further research on no MS-factors is required to predict better elevated CIMT.


Background
Cardiovascular disease (CVD) is the leading cause of mortality and morbidity in subjects with T2DM worldwide, and physicians call coronary equivalent to that subject with diabetes mellitus without coronary disease due to the high burden of disease. Subclinical cardiovascular disease (SCVD) determination, through Carotid intima-media (CIMT), coronary artery calcium (CAC), and presence of carotid plaque (CP), is associated with overt CVD, and this structural surrogate helps establish preventive strategies in primary care. Although CIMT has a lower predictive ability to detect overt CVD than CAC and CP (1), it might be considered the earliest structural change compared to the other subclinical parameters.
Metabolic syndrome (MS) is the aggregation of factors that include three of ve components: waist circumference, glucose, HDL-cholesterol, triglycerides, and blood pressure (2). There are multiple de nitions, varying by the cut-off points or one factor's predominance over another (2)(3)(4). Both T2DM and MS are associated with an increased prevalence of SCVD and overt CVD (5). Furthermore, the reduction or aggregation of factors does not necessarily determine lower o higher cardiometabolic risk. Gurka et al. (7) designed a continuous score for the severity of metabolic syndrome speci c to sex and ethnic group. According to factor analysis models, the metabolic syndrome severity score (MSSS) offers the advantage of converting those dichotomous variables (present or absent) into continuous numerical variables. This severity score has shown a good correlation with risk markers such as insulin, adiponectin (8), and predictive value for cardiovascular risk in T2DM subjects (9), and chronic kidney disease (10).
For these reasons, it is critical to determine whether MSSS can predict the SCVD through the CIMT measurement and to clarify if this method could be used as an early tool for preclinical CVD. The absence of SCVD seems to attenuate the risk of overt CVD in subjects with T2DM and MS (6). Thus, we propose evaluating the accuracy of MSSS and other metabolic syndrome de nitions to detect increased carotid intima-media thickness in Peruvian Type 2 diabetes mellitus subjects without CVD attending to a healthcare network social security system.

Study design.
The study was conducted as a cross-sectional design. All subjects gave written informed consent, and the study procedure followed the Helsinsky Declaration's agreements and was approved by the institutional ethics committee.

Patient
A total of 184 subjects with Diabetes mellitus were recruited from the outpatient clinic of the Edgardo Rebagliati Martins Hospital (HNERM) and the Diabetes and Hypertension Center (CEDHI) of the Social security Health System between January 2019 and January 2020. We included subjects older than 20 years and diagnosis of DM based on the ADA criteria. Those patients with a history of coronary artery disease, stroke, peripheral revascularization, or major amputation due to peripheral arterial disease were excluded. Likewise, those with DM1, LADA, NYHA III or IV heart failure, Child B or C liver disease, glomerular ltration rate < 60 ml/min, acute or previous disease up to a range of 6 months, surgery & radiotherapy at the cervical level and use of immunosuppressants or corticosteroids.

Clinical and biochemical parameters
Independent, trained medical personnel evaluated and collected data patients. Vital functions and anthropometric measurements were taken according to the established protocol. Waist circumference was measured with a tape scale according to the International Diabetes Federation (IDF) recommendations. A questionnaire was lled out directly from the patient regarding pathological history, lifestyle, comorbidities, and pharmacological treatment.
We obtained fasting blood samples for biochemical analyzes that included glucose, HbA1c, and lipid pro le. The process was according to the central laboratory's routine method.
We de ned Metabolic syndrome according to ATP III and harmonized criteria and use the cut-of for abdominal perimeter obtained by Aschner in the harmonized MS criteria. We calculated MSSS (7) and non-glucose MSSS, according to Gurka et al (9), as Z-score through a spreadsheet.

Measurement of MICT
We measured CIMT by a single trained physician independent of data collection and processing using a Toshiba Nemio XG ultrasound scanner. We used a B-mode carotid ultrasound with a 7.5 MHz linear transducer to assess the CIMT with standard insonation parameters. Patients were examined supine with the neck hyperextended and identi ed three segments in the anterior and posterior plane: 1 cm distal to the common carotid proximal to the carotid bifurcation, the bifurcation itself, 1 cm proximal to the internal carotid artery. We obtained the maximum average CIMT value from the far wall of both the right and left common carotid arteries and excluded CIMT ≥ 1.5 mm

Statistical analysis
We presented categorical data as absolute and relative frequencies and consecutive data as mean and standard deviation or medians with interquartile range according to normal distribution. We described CIMT according to sex and age.
Also, we show the correlation between each SM de nition with CIMT through Pearson's coe cient and its scatter charts. A value of p < 0.05 was considered signi cant.
We performed area Under the Receiver Operator Curve (AUROC) of each MS de nition to predict CIMT ≥ 0.86 mm, and we calculated sensitivity and speci city according to the optimal cut-off value for each MS de nitions.
We made a bivariate analysis for related and no related MS factors for CIMT and a regression model A adjusted to age and sex, and a regression model B adjusted to age, sex, and HbA1c.

Results
This study included 184 subjects with T2DM 29% were men with a mean age of 61.5 ± 10.5 years old. Median CIMT was 0.8 mm (0.7-0.9), and the prevalence of elevated CIMT, de ned by Rizzo (≥ 0.86 mm), was 29.4%. CIMT tends to increase according to age in both sexes (Fig. 1) Only non-glucose MSSS showed a correlation with CIMT. (r = 0.19, p < 0.01). (Fig. 2). Non-glucose MSSS obtained the best AUROC 0.61 (0.52-0.70) compared with the other de nitions. (Fig. 3). MSSS and Harmonized SM criteria included 0.5 in their con dence interval. A Z-score of -1.0 was the best cut-of point of non-glucose MSSS to predict elevated CIMT. This value had 57.4% and 60% of sensitivity and speci city, respectively. (Table 2)  Only systolic pressure was associated with CIMT (PR 1.01 CI95% 1.01 to 1.03; p < 0.000) of the metabolic syndrome factors. Age, male sex, and the use of hypertensive drugs also showed association with CIMT (Table 3).

Discussion
MSSS is a prediction model developed from the National Health and Nutrition Examination Survey (NHANES) database. It must be convenient to validate outcome studies in different populations such as Latin American subjects to answer whether this prediction model can predict subclinical atherosclerosis as an early marker of CV events. In the present study, only the non-glucose MSSS de nition showed the best diagnostic discrimination for elevated CIMT and presented a weak performance. The original MSSS did not show any discrimination in this population.
MSSS performed a signi cant association with future CVD among type 2 DM subjects from the Atherosclerosis Risk in Communities study (ARIC) (9). This cohort study found that MSSS and nonglucose MSSS at baseline were associated with incident CVD HR 1.29 ( 95% CI 1.21-1.39; p < 0.001).and HR 1.42 (95% CI 1.24-1.62; p < 0.001), respectively. This outcome study explains our better predictive ability of non-glucose MSSS to detect elevated CIMT than the original MSSS.
We did not nd studies related to the association between MSSS and subclinical atherosclerosis as a surrogate marker of cardiovascular risk. Gurka et al. showed that MSSS has an excellent predictive performance of cardiovascular risk if glucose is not considered (9) or even when fasting values is replaced for non-fasting values into this model (11). The independent association of non-glucose-modi ed MSSS with cardiovascular risk might be explained for a possible weakness of glycemic status for detecting CIMT. Other risk factors as age, male sex, high-density-lipoprotein cholesterol, and total cholesterol also in uence their development (12). However, glycemic parameters as post-prandial glucose (13) and glycemic variability are associated as independent risk factors for CIMT (14). The oxidative stress related to glucose uctuations could explain this association. So glycemic control is an essential therapeutic intervention to decrease the subclinical atherosclerosis process (15).
DM and MS are associated with atherosclerotic ndings as coronary calcium content, abdominal aortic calci cation, elevated CIMT, and abnormal ankle-brachial index (ABI). These parameters, alone or combined, with greater cardiovascular risk rate (16). CIMT is one of the earliest surrogates of subclinical cardiovascular disease, and its measurement is easy to use even in-o ce settings, although its interpretation is operator dependent. Additionally, identifying biomarkers with the ability to predict early structural change, such as CIMT, should be an essential tool for modifying the atherosclerosis process's natural history as a primary prevention strategy for CVD (17)(18)(19)(20)(21).
As a weakness, the study stopped due to the SARS CoV 2 pandemic, and the original sample size did not complete, although the trend would not change in light of the results. Another limitation was that cardiovascular disease background was referred for the subjects without any assessment to rule out subclinical coronary disease. We selected the subjects by non-random sampling from a reference hospital and a primary care center and did not extrapolate beyond the study population. Finally, the small number of males and a wide age range could miss estimating the true prevalence of altered CIMT. As a strength, we evaluated the carotid intima-media thickness with the appropriate technique by on a highresolution ultrasound and a trained operator.

Conclusions
Our results show that non-glucose modi ed MSSS had the greatest accuracy for detecting CIMT compared with MS-ATP III, MS-harmonizing criteria, and original MSSS. Further studies are necessary to evaluate ultrasonography's clinical utility as an early intervention to reduce the risk of CIMT progression. Also, to evaluate the role of metabolic syndrome in the development of subclinical atherosclerosis.

Declarations
Ethics approval and consent to participate: Informed consent to participate in the study wasobtained from participants. The informed consent document was approved by the institutional ethics committee.

Consent for publication:
Not applicable for that section.
Availability of data and materials: The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.