Soluble ST2, a biomarker of fibrosis, is associated with multiple risk factors, chronic diseases and total mortality in the OPERA study

Figures & data

Figure 1. OPERA – Oulu Project Elucidating Risk of Atherosclerosis study outline.

Table 1. Main characteristics of the baseline and follow-up study population.

	Control cohort	Hypertensive cohort
Age (years)
Baseline	51.4 (6.0)	51.2 (5.9)
Follow-up	71.7 (5.5)	71.7 (5.3)
Female (%)
Baseline	50.8 (n = 267)	49.7 (n = 258)
Follow-up	56.1 (n = 174)	50.0 (n = 145)
BMI (kg/m^2)
Baseline	26.4 (4.0)	29.0 (4.9)^a
Follow-up	28.1 (4.7)	30.2 (5.2)^a
Waist circumference (cm)
Baseline	87.2 (12.0)	94.0 (13.5)^a
Follow-up	93.9 (13.6)	100.4 (13.4)^a
Systolic BP (mmHg)
Baseline	141.0 (21.0)	155.0 (21.0)^a
Follow-up	137.6 (22.5)	138.9 (21.7)
Diastolic BP (mmHg)
Baseline	85.0 (12.0)	94.0 (11.0)^a
Follow-up	71.6 (10.6)	72.9 (10.3)
Smoking habit (pack years)
Baseline	12.4 (14.3)	11.5 (13.7)^a
Follow-up	10.6 (17.2)	10.30 (17.4)
Alcohol consumption (g/week)
Baseline	56.0 (82.0)	68.0 (103.0)
Follow-up	40.1 (70.9)	38.3 (66.7)
ST2 (log transformed)
Baseline	–	–
Follow-up	1.3 (0.16)	1.3 (0.18)
Diabetes status (%)
Baseline	4.2	13.7^a
Follow-up	13.8	22.2^a
Creatinine (µmol/L)
Baseline	79.9 (12.8)	85.4 (42.9)^a
Follow-up	69.3 (18.5)	77.7 (37.0)^a
eGFR (mL/min/1.73 m^2)
Baseline	83.8 (16.4)	82.2 (17.4)^a
Follow-up	84.0 (12.7)	79.5 (17.2)^a
Total cholesterol (mmol/L)
Baseline	5.6 (1.1)	5.8 (1.0)
Follow-up	4.9 (1.0)	4.5 (1.0)^a

BMI: body mass index; systolic BP: systolic blood pressure, mean of 2 and 3 measurements; diastolic BP: diastolic blood pressure, mean of 2 and 3 measurements; eGFR: estimated glomerular filtration rate.

Values given in means or percentages (SD). Diabetes status (no = 0, yes = 1). ^aStatistical significant difference (p < 0.05) between cohorts.

Table 2. Baseline classical and metabolic risk predictors of follow-up plasma ST2 levels in univariate and multivariate regression models.

	Univariate model		Multivariate model
	p Value	Beta coefficient	p Value	Beta coefficient
Gender (female = 0, male = 1)	<.001	–0.197	.174	–0.083
Diabetes status (no = 0, yes = 1)	.024	0.092	.135	0.065
QUICKI	<.001	–0.172	.829	–0.012
Waist circumference (cm)	<.001	0.227	.465	0.046
Hepatosteatosis	<.001	0.185	.057	0.090
Smoking habit (pack years)	<.001	0.163	.012	0.113
Alcohol consumption (g/week)	.018	0.097	.328	–0.045
Systolic BP (mmHg)	<.001	0.153	.163	0.063
Creatinine (µmol/L)	.001	0.136	.148	0.067
Total cholesterol (mmol/L)	.033	0.087	.603	0.023
HDL-cholesterol (mmol/L)	.002	–0.129	.563	0.031
Triglycerides (mmol/L)	<.001	0.168	.636	0.026
hsCRP (mg/L)	.003	0.123	.448	0.035
Antihypertensive medication (no = 0, yes = 1)	.003	0.122	.301	0.045
Fasting glucose (mmol/L)	.001	0.139
Fasting insulin (mmol/L)	<.001	0.150
BMI (kg/m^2)	<.001	0.150
ALAT (U/L)	<.001	0.173
Diastolic BP (mmHg)	<.001	0.161
LDL-cholesterol (mmol/L)	.020	0.095
Carotid intima-media thickness (mm)	.045	0.082

QUICKI: quantitative insulin sensitivity check-index; systolic BP: systolic blood pressure, mean of 2 and 3 measurements; HDL: high-density lipoprotein; hsCRP: high-sensitivity C-reactive protein (log transformed); BMI: body mass index; P-ALAT: plasma alanine aminotransferase; diastolic BP: diastolic blood pressure, mean of 2 and 3 measurements; LDL: low-density lipoprotein.

Shaded values indicate a significant p-value of the parameter in both univariate and multivariate models.

Table 3. Association of classical and metabolic risk factors with plasma ST2 levels at follow-up examination, in univariate and multivariate regression models.

	Univariate model		Multivariate model
	p Value	Beta coefficient	p Value	Beta coefficient
Gender (female = 0, male = 1)	<.001	–0.197	<.001	–0.215
Diabetes status (no = 0, yes = 1)	<.001	0.211	<.001	0.159
QUICKI	<.001	–0.227	.006	–0.141
Waist circumference (cm)	<.001	0.182	.067	–0.102
Smoking habit (pack years)	.001	0.141	.119	0.065
Alcohol consumption (g/week)	.042	0.083	.702	–0.016
Exercise during free time^a	<.001	–0.147	.074	–0.074
Creatinine (µmol/L)	<.001	0.154	.067	0.075
Total cholesterol (mmol/L)	.014	–0.101	.516	–0.030
HDL-cholesterol (mmol/L)	.047	–0.081	.013	0.122
hsCRP (mg/L)	<.001	0.165	<.001	0.160
P-ALAT (U/I)	<.001	0.150	.042	0.083
Antihypertensive medication (no = 0, yes = 1)	.003	0.122	.452	0.032
P-glucose (mmol/L)	.002	0.240
S-insulin, mU/L	<.001	0.182
BMI (kg/m^2)	<.013	0.102
eGFR (mL/min/1.73 m^2)	.010	–0.105

QUICKI: quantitative insulin sensitivity check-index; HDL: high-density lipoprotein; hsCRP: high-sensitivity C-reactive protein (log-transformed); P-ALAT: plasma alanine aminotransferase concentration; P-glucose: plasma glucose; S-insulin: serum insulin; BMI: body mass index; eGFR: estimated glomerular filtration rate.

^a Exercise during free time: no or irregular, regular, heavy regular.

Shaded values indicate a significant p-value of the parameter in both univariate and multivariate models.

Figure 2. The association of follow-up glucose tolerance groups with plasma levels of sST2.

Table 4. Association of plasma sST2 (log-transformed) levels with follow-up diseases in multivariate model.

	Yes	No	Adjusted p Value
Metabolic syndrome	1.34 (0.17) (n = 370)	1.32 (0.17) (n = 228)	.063^a
Sleep apnea	1.40 (0.17) (n = 48)	1.32 (0.17) (n = 550)	.053^a
Coronary artery disease	1.38 (0.18) (n = 127)	1.32 (0.17) (n = 471)	.018^b
Any evidence of CVD^d	1.37 (0.18) (n = 189)	1.31 (0.17) (n = 409)	.031^b
Cancer	1.37 (0.20) (n = 110)	1.32 (0.17) (n = 488)	.021^a
Cognitive impairment (MMSE)^e	1.39 (0.14) (n = 44)	1.33 (0.17) (n = 554)	.046^c

The results are given as means (SD). The ‘yes’ group consisted of subjects whose scores were lower than or equal to 24 while the ‘no’ group comprised the rest of the subjects (MMSE higher than 24).

^a Multivariate model includes sex, quick-index, plasma HDL-cholesterol, hsCRP, ALAT and diabetes status (not in metabolic syndrome).

^b Adjusted also with age, LDL-cholesterol, smoking (pack-years) and hypertension status.

^c Multivariate model includes age, sex and hsCRP.

^d CVD: cardiovascular disease = coronary artery disease, cerebrovascular disease or peripheral vascular disease.

^e MMSE: Mini-Mental State Examination.

Figure 3. sST2 tertile groups and the risk of death.