This study is a long-term register study of the Hämeenlinna Metabolic Syndrome Research Program. The study subjects have been described earlier [8]. In short, the study subjects were 120 men with MetS according to the National Cholesterol Education Program (NCEP) Adult Treatment Panel III criteria [9] aged from 30 to 65 years from the Hämeenlinna region (population 170 000). The study subjects were referred from private and public consultations in primary and secondary health care. In the initial study, in addition to the traditional baseline characteristics, we measured the subjects’ arterial elasticity, hs-CRP and OxLDL concentrations and calculated the 10-year risk of CVD events and mortality using SCORE and FINRISK score models. The FINRISK model uses the subject’s age, sex, smoking status, systolic blood pressure, total cholesterol, high-density lipoprotein cholesterol (HDL-C), diabetes status and possible AMI or stroke of subject’s parents to predict the risk of AMI or stroke during the next 10 years [3]. The SCORE risk model uses the subject’s age, sex, smoking status, systolic blood pressure, total cholesterol or total cholesterol/HDL-C ratio and population-based risk to predict the risk of fatal cardiovascular event during the next 10 years [2].

The incidence of composite endpoints was assessed in three subject groups for both risk score models: low, medium and high risk. For FINRISK, the 10-year risk for fatal or non-fatal AMI or stroke was < 5% in low-risk group, 5-14.9% in medium-risk group and ≥ 15% in high-risk group. For SCORE, the risk of CVD death for the groups were < 3%, 3-4.9% and ≥ 5%, respectively.

The subjects filled a structured questionnaire including their smoking status and typical weekly amount, type and mode of physical activity at baseline. The information on the subjects’ cardiovascular disease status and mortality was retrieved from the region’s electronic medical record (EMR) on 25. March, 2021.

The study followed the ethical principles of the Declaration of Helsinki and each study subject gave a written informed consent. The original study protocol was approved by the Research Ethics Committee of Kanta-Häme Hospital District. The subject recruitment and baseline characteristics were performed 15.6.2004–10.11.2005 so the minimum follow-up data of 15 years was available from all subjects. The follow-up data of ≥ 15 years was collected in a registry setting and the follow-up study was approved by the Ethics Committee of Tampere University Hospital District (no. R19057).

Blood samples were drawn after a 12-hour overnight fast with a minimum of 10-minute rest. The serum total cholesterol, low-density lipoprotein cholesterol (LDL-C), HDL-C, triglyserides, OxLDL and glycated hemoglobin (HbA1c) were measured and analyzed in hospital laboratory, which practices strict internal quality control and national external quality assurance program (Labquality Oy, Helsinki, Finland) with daily and monthly control sampling. Cutoffs for OxLDL were following: <60 U/L for low risk, ≥ 60 and < 70 U/L for moderate risk and ≥ 70 U/L for high risk. These robust cutoffs for general guidance are provided by commercial laboratories and they are based on a study of 1,889 apparently healthy individuals without MetS and a reference range study using specimens from 704 apparently healthy volunteers [10].

The plasma concentration of hs-CRP test was analyzed according to the validated and accredited method in the Vita Laboratories, Helsinki, Finland. In brief, the analysis of hs-CRP was carried out with a commercially available Bekman Coulter CRP latex reagents used with an Olympus AU620 analyzer (Beckman Coulter, Brea, CA, USA). The measurement range of this immunoturbidometric test was hs-CRP level between 0.2 and 160 mg/l, and the reference value was used as a CRP level 3 mg/l. A hs-CRP-level under 1.0 mg/l indicates a low cardiovascular risk, a hs-CRP-level 1.0–3.0 mg/l indicates a moderate cardiovascular risk, and a hs-CRP-level above 3.0 mg/l indicates a high cardiovascular risk [11]. After assay, the hs-CRP results were validated and released through controlled validation process.

The arterial elasticity was measured after a minimum of 10-minute rest with a non-invasive arterial tonometer (HDI/PulseWave™ CR-2000, Hypertension Diagnostics, Eagan, MN, USA) which records the radial artery pulse wave. The tonometer utilizes a modified Windkessel method [12] to report two variables as a mean of five most even pulse waves measured during the recording: capacitive elasticity of large arteries (C1) and the reflective elasticity of small arteries (C2). The C1 depicts the elastic properties of large arteries such as aorta or coronary arteries and the C2 depicts the endothelial function of the microvascular circulation [13]. The measuring technique has been shown to correlate with invasive measures of arterial elasticity with high reproducibility [14, 15]. The C1 and C2 were measured four times for each subject to obtain mean large and small arterial elasticity results. The subjects’ arterial elasticity was graded according to the CR-2000 normal values provided by the manufacturer (Table 1).

The primary outcome of the study was a composite of (I) the first occurrence of type 1 AMI or cerebral stroke, (II) diagnosis of symptomatic coronary or peripheral artery disease (PAD) diagnosed with invasive angiography, (III) the first occurrence of coronary or peripheral revascularization, (IV) amputation due to PAD, (V) cardiovascular death and (VI) non-cardiovascular death. Only one endpoint event was accepted for each hospitalization i.e., a subject with AMI leading to coronary angiography and percutaneous coronary intervention (PCI) would only be registered as having one endpoint instead of three. Transient ischemic attack (TIA) was not accepted as an endpoint since the diagnosis relies strongly on clinical suspicion in the absence of other plausible causes. A single cardiologist (HS) individually assessed all endpoints and was blinded to the baseline characteristics of the subjects.

Statistical analysis was done with IBM SPSS Statistics (Version: 29.0.1.0, 2023). Cumulative survival curves were generated with the use of the Kaplan-Meier method and difference of survival between groups was tested with the log rank test. The p-value of < 0.05 was held statistically significant.