Increased serum galectin-3 level is associated with endothelial dysfunction and cardiovascular events in patients with hypertension

Abstract

Background

Endothelial dysfunction can lead to various harmful cardiovascular complications. The importance of galectin-3 (Gal-3) has been proposed in some cardiac diseases related to chronic inflammation. However, its role in hypertension-induced endothelial dysfunction remains unclear.

Methods

We enrolled 120 patients with hypertension, assessed their baseline characteristics, and monitored their 7-year cardiovascular outcomes. We performed an enzyme-linked immunosorbent assay to measure serum Gal-3 levels. The vascular reactivity index (VRI) was examined by digital thermal monitoring. Patients with VRI <1.0, 1.0 to <2.0, and ≥2.0 were defined as having poor, intermediate, and good vascular reactivity, respectively.

Results

Among the recruited patients, 12 had poor vascular reactivity, 57 had intermediate vascular reactivity, and 51 had good vascular reactivity. Older age, higher total cholesterol levels, higher low-density lipoprotein cholesterol levels, lower estimated glomerular filtration rate, and higher Gal-3 levels were associated with poor endothelial dysfunction. Multivariate linear regression analysis showed that age and Gal-3 levels were correlated with VRI. During the 7-year follow-up period, patients with higher Gal-3 levels had more cardiovascular events.

Conclusions

Higher Gal-3 levels are associated with endothelial dysfunction and unfavorable cardiovascular outcomes in patients with hypertension, suggesting its potential role in the hypertension-induced endothelial dysfunction.

Highlights

• •Endothelial dysfunction is a hallmark of hypertension, and promotes the development of cardiovascular complications.
• •Higher serum Gal-3 level correlates with worse endothelial function and unfavorable cardiovascular prognosis.
• •Gal-3 could be a potential biomarker for predicting cardiovascular outcomes.

1. Introduction

Hypertension is a severe medical problem which substantially increases the risk of various cardiovascular (CV) complications, including atherosclerosis, stroke, congestive heart failure, and peripheral vascular disease [1]. Over the past decades, tremendous efforts have been made to control blood pressure (BP), however, only 14 % of hypertensive patients worldwide have reached their BP target according to data from the World Health Organization. Moreover, the mechanisms through which hypertension causes vascular damage remain to be elucidated.

Endothelial dysfunction (ED) is one of the significant hallmarks of hypertension and can be further aggravated by increased BP [2,3]. The integrity of endothelium is maintained by several mediators [3], and could be jeopardized by various damaging signals, such as oxidative stress and abnormal vascular shear stress [4,5]. Functional impairment of the endothelium leads to disruption of endothelium-derived vasodilation, smooth muscle cell proliferation, leukocyte adhesion, and migration, followed by the development or aggravation of atherosclerotic cardiovascular diseases (CVD) [6,7]. Endothelial dysfunction also linked to the new-onset diabetes in hypertensive patients [8]. Importantly, ED is not confined to a peripheral area but is a systemic disorder [4]. Thus, detecting and preventing ED in the early stage is critical. Over the past few years, galectins are found to be associated with oxidative stress and CVD and have become an emerging research interest [[9], [10], [11]].

Galectins, a family of proteins that bind to β-galactose residues through a carbohydrate recognition domain, regulate various essential biological functions extracellularly and intracellularly [[12], [13], [14]]. Galectin-3 (Gal-3) is a 29–35-kDa protein with a unique chimera structure [15], and express in many cells in different quantities [16]. Although not cardiac-specific, it has been studied for putative roles in CV-related physiological and pathological processes over the past few years. In patients with heart failure, elevated serum Gal-3 levels are significantly associated with worse outcomes [17,18]. And the levels of Gal-3 have been suggested to serve as a biomarker for monitoring the heart failure status [[17], [18], [19], [20]]. In hypertension, Gal-3 is found to promote cardiac remodeling and myocardial fibrosis [21]. The inhibition of Gal-3 seems to attenuate hypertension-related complications [22]. Based on this evidence, we propose that Gal-3 is involved in hypertension-related ED, which leads to harmful CV outcomes.

Although the importance of Gal-3 has been addressed in heart failure, its role in endothelial function and prognostic value have not yet been elucidated. In this study, we investigated the correlation between Gal-3 and endothelial function in patients with hypertension. Moreover, we explored whether Gal-3 could serve as a biomarker of CV outcomes in these individuals.

2. Methods

2.1. Study population and baseline clinical evaluation

Patients with hypertension were enrolled in the CV clinic of Hualien Tzu Chi Hospital between May 2016 and December 2016. We excluded patients with acute infections, acute myocardial infarction, malignancy, or the usage of sodium glucose cotransporter 2 inhibitor or glucagon-like peptide 1 receptor agonist. Patients who refused to provide informed consent were also excluded from this study. In total, 120 patients with hypertension were enrolled in this study. After recruitment, the recruited patients’ clinical characteristics were collected from medical records. Patients with >50 % stenosis in any segment of the coronary arteries were defined as having coronary artery disease (CAD). Patients with fasting plasma glucose levels ≥126 mg/dL or those taking antihyperglycemic medications were defined as having diabetes mellitus (DM). The definition of hypertension was either (1) patients had high BP according to the Eighth Joint National Committee (JNC 8) guideline, or (2) patients with a diagnosis of hypertension and under antihypertensive agents treatment. All hypertensive patients enrolled in this study received antihypertensive medications during the study period. All participants received venipuncture at fasting status for laboratory analysis. Their endothelial function was evaluated by vascular reactivity index using non-invasive digital thermal monitoring. The protocol for this study was approved by the Research Ethics Committee at Hualien Tzu Chi Hospital, Buddhist Tzu Chi Medical Foundation (IRB 104-27-B). At the time of the interview, all participants signed informed consent forms and written agreements stating they agreed to participate in this study.

2.2. Digital thermal monitoring

To evaluate endothelial function, digital thermal monitoring was performed [[23], [24], [25], [26]]. For this, all patients fasted overnight and avoided the use of tobacco, alcohol, and caffeine. Endothelial function was measured using digital thermal monitoring using an FDA-approved device, VENDYS-II (Endothelix, Inc., Houston, TX, USA). During the measurement, BP cuffs were placed on the bilateral upper arms, and skin temperature sensors were placed on both index fingers. After stabilizing for 3 min, we inflated the BP cuff to 50 mmHg greater than the patient's systolic BP, maintained it for 5 min, and then rapidly deflated the cuff. After releasing the cuff, reactive hyperemia occurred, and the temperature of the fingertips increased. We measured the maximum difference between the observed temperature rebound curve and the zero-reactivity curve during the reactive hyperemia period using VENDYS to determine the VRI. Patients with VRI <1.0, 1.0 to <2.0, and ≥2.0 were defined as having poor, intermediate, and good vascular reactivity, respectively.

2.3. Blood analysis

After overnight fasting, 5 mL of peripheral blood was collected from each patient. The sample was centrifuged at 3,000 g for 10 min, and the serum was stored at 4 °C for subsequent analysis. To measure the levels of fasting glucose, albumin, blood urea nitrogen, creatinine, total cholesterol, triglycerides, high-density lipoprotein cholesterol (HDL-C), and low-density lipoprotein cholesterol (LDL-C), we used an autoanalyzer (Siemens Advia 1800; Siemens Healthcare, Henkestr, Germany) to obtain the results from each sample. To determine the protein levels of Gal-3, we performed an enzyme-linked immunosorbent assay using a commercial kit (RayBiotech, Peachtree Corners, Georgia, USA) [27]. The estimated glomerular filtration rate (eGFR) was calculated using the Chronic Kidney Disease Epidemiology Collaboration equation.

2.4. Follow-up and outcomes

The patients were followed up regularly in person at the CV clinic of Hualien Tzu Chi Hospital. In total, 104 patients with hypertension were followed up in this study. Surveillance of any CV events and CV hospital admissions was performed in the first 7 years after enrollment according to the medical records. CV events were defined as cardiovascular death, myocardial infarction, heart failure, ischemic stroke and hemorrhagic stroke. Event-free survival was defined as the duration between enrollment evaluation and either the occurrence of the prespecified endpoint or the end of the study follow-up.

2.5. Statistical analysis

For categorical variables, we counted the number of patients in each specific category, and then checked the statistical significance among these profiles by Chi-square test. For the continuous variables, Kolmogorov–Smirnov test was performed in the first place for examining whether these sets of data were come from normal distribution. Normally distributed variables are presented as means ± standard deviations, and the means between these groups were compared by two-tailed independent Student's t-test or one-way ANOVA. The non-normal variables are presented as medians and interquartile ranges (IQRs), and the comparisons between groups were carried out by Mann–Whitney U test or Kruskal–Wallis test. To exam the correlation of clinical variables and VRI, simple linear regression and multivariate regression analysis were performed. To further exam the association between Gal-3 levels and vascular reactivity dysfunction, univariate and multivariate logistic regression analyses were done. For the CV events occurred during follow-up period, multivariate Cox regression models was used for identifying the variables that are associated with the occurrence of CV events. After dividing patients into two groups according to their serum Gal-3 level (higher or lower than the mean value 12.87 ng/mL), the Kaplan–Meier method was used for analyzing their event free survival, and the log-rank test was performed for comparing the cumulative portion of patients who were free from CV events. p-values <0.05 were judged as statistical significance. All statistical analyses were performed using Statistical Package for the Social Sciences (version 19.0; IBM Corp., Armonk, NY, USA).

3. Results

3.1. Baseline characteristics

In this study, 120 patients with hypertension were prospectively enrolled. Among the participants, 52 (43 %) had DM, 89 (74 %) had CAD, and 25 (21 %) were current smokers. The patients were then grouped according to their vascular reactivity, and 51 patients (43 %) had good vascular reactivity, 57 patients (47 %) had intermediate vascular reactivity, and 12 patients (10 %) had poor vascular reactivity. As shown in Table 1, the basic clinical characteristics among these three subgroups were similar, including sex distribution, body mass index, use of cigarettes, presence of DM or CAD, BP control, and types of antihypertensive medications.

Table 1.

Clinical characteristics according to different vascular reactivity indexes by digital thermal monitoring of the 120 hypertensive patients.

Characteristics	All participants (n = 120)	Good vascular reactivity (n = 51)	Intermediate vascular reactivity (n = 57)	Poor vascular reactivity (n = 12)	p value
Age (years)	63.37 ± 7.44	61.66 ± 6.58	63.77 ± 7.79	68.68 ± 6.98	0.010∗
Height (cm)	164.48 ± 7.38	163.45 ± 8.04	165.31 ± 7.08	164.92 ± 5.66	0.418
Body weight (kg)	71.60 ± 11.63	70.33 ± 9.53	73.65 ± 12.91	67.23 ± 12.24	0.130
Body mass index (kg/m2)	24.43 ± 3.69	26.36 ± 3.42	26.87 ± 3.85	24.62 ± 3.73	0.156
Vascular reactivity index	1.86 ± 0.63	2.41 ± 0.33	1.63 ± 0.25	0.58 ± 0.22	<0.001∗
Systolic blood pressure (mmHg)	137.53 ± 18.17	140.27 ± 17.99	134.11 ± 17.57	142.08 ± 20.26	0.139
Diastolic blood pressure (mmHg)	80.37 ± 10.81	82.84 ± 10.54	78.11 ± 10.91	80.58 ± 9.96	0.074
Mean arterial pressure (mmHg)	99.42 ± 12.16	101.99 ± 11.84	76.77 ± 12.07	101.08 ± 12.27	0.073
Total cholesterol (mg/dL)	164.33 ± 38.36	160.35 ± 33.43	161.58 ± 37.48	194.33 ± 50.89	0.015∗
Triglyceride (mg/dL)	129.00 (98.25–187.75)	122.00 (101.00–182.00)	127.00 (94.50–191.00)	153.50 (90.50–214.75)	0.948
HDL-C (mg/dL)	46.00 (39.00–55.75)	46.00 (40.00–55.00)	44.00 (37.50–56.00)	49.00 (39.75–56.25)	0.334
LDL-C (mg/dL)	92.93 ± 27.98	88.45 ± 21.66	92.32 ± 26.98	114.92 ± 44.79	0.011∗
Fasting glucose (mg/dL)	108.00 (91.25–135.75)	111.00 (92.00–139.00)	105.00 (91.50–137.50)	103.00 (88.50–125.50)	0.523
Blood urea nitrogen (mg/dL)	17.00 (14.00–19.00)	16.00 (13.00–18.00)	17.00 (14.00–20.50)	18.50 (13.25–24.50)	0.197
Creatinine (mg/dL)	1.00 (0.80–1.10)	0.90 (0.80–1.10)	1.00 (0.90–1.10)	1.00 (0.90–1.08)	0.283
eGFR (mL/min)	83.01 ± 23.34	88.79 ± 25.87	79.95 ± 20.98	73.04 ± 17.32	0.042∗
Galectin-3 (ng/mL)	13.11 ± 5.33	10.76 ± 4.06	13.91 ± 4.64	19.31 ± 7.22	<0.001∗
Male, n (%)	100 (83.3)	40 (78.4)	50 (87.7)	10 (83.3)	0.433
Diabetes mellitus, n (%)	52 (43.3)	23 (45.1)	22 (38.6)	7 (58.3)	0.431
Coronary artery disease, n (%)	89 (74.2)	38 (74.5)	43 (75.4)	8 (66.7)	0.817
Smoking, n (%)	25 (20.8)	14 (27.5)	9 (15.8)	2 (16.7)	0.307
ACE inhibitor or ARB use, n (%)	78 (65.0)	34 (66.1)	37 (64.9)	7 (58.3)	0.862
β-blocker use, n (%)	75 (62.5)	28 (54.9)	39 (68.4)	8 (66.7)	0.333
CCB use, n (%)	49 (40.8)	24 (27.1)	20 (35.1)	5 (41.7)	0.449
Statin use, n (%)	96 (80.0)	38 (74.5)	48 (84.2)	10 (83.3)	0.433
Fibrate use, n (%)	8 (6.7)	4 (7.8)	3 (5.3)	1 (8.3)	0.841

Values for continuous variables are given as means ± standard deviation and tested by one-way analysis of variance; variables not normally distributed are given as medians and interquartile range and tested by Kruskal-Wallis analysis; values are presented as number (%) and analysis after analysis by the chi-square test.

HDL-C, high-density lipoprotein cholesterol; LDL-C, low-density lipoprotein cholesterol; eGFR, estimated glomerular filtration rate; ACE, angiotensin-converting enzyme; ARB, angiotensin-receptor blocker; CCB, calcium-channel blocker.

∗p < 0.05 was considered statistically significant after the Kruskal-Wallis or one-way variance analysis.

3.2. Gal-3 is negatively correlated with the VRI and is independently associated with ED

In our study, we identified that the proportions of patients with older age (p = 0.010), higher serum levels of total cholesterol (p = 0.015) and LDL-C (p = 0.011), and lower eGFR levels (p = 0.042) were significantly higher in the poor and intermediate VRI groups than in the good VRI group (Table 1). Furthermore, we discovered a novel target, Gal-3 (p < 0.001), which was significantly higher in the poor and intermediate VRI groups than in the good VRI group. Participants with poor endothelial function had significantly higher serum Gal-3 levels (Table 1).

To further validate the correlation between clinical variables and VRI, we performed simple and multivariate linear regression analyses (Table 2). Age (r = −0.279, p = 0.002) and serum levels of total cholesterol (r = −0.239, p = 0.009), LDL-C (r = −0.304, p = 0.001), and Gal-3 (r = −0.512, p < 0.001) were negatively correlated with VRI, whereas eGFR (r = 0.200, p = 0.028) was positively correlated with VRI. VRI was independently correlated with age (β = −0.224, adjusted R² change = 0.044, p = 0.004) and serum Gal-3 levels (β = −0.387, adjusted R² change = 0.255, p < 0.001) on multivariate forward stepwise linear regression analysis. To confirm the association between Gal-3 and vascular reactivity and avoid potential confounding factors in our analysis, we performed multivariate logistic regression analysis. We used three models and adjusted for age, sex, body mass index, BP, fasting glucose, eGFR, lipid profile, anti-hypertensive drugs, and anti-lipid drugs. Specifically, higher Gal-3 levels were positively and independently associated with poor vascular reactivity (odds ratio (OR) = 1.222, 95 % confidence interval (CI) = 1.079–1.384, p = 0.002) and vascular reactivity dysfunction (OR = 1.373, 95 % CI = 1.095–1.723; p = 0.006) (Table 3). These results suggest that Gal-3 levels are an independent marker of endothelial function in patients with hypertension.

Table 2.

Correlation of vascular reactivity index levels and clinical variables by simple or multivariable linear regression analyses among 120 hypertensive patients.

Variables	Vascular reactivity index
	Simple regression		Multivariable regression
	r	p value	Beta	Adjusted R2 change	p value
Male	0.076	0.410	–	–	–
Diabetes mellitus	0.036	0.699	–	–	–
Coronary artery disease	0.124	0.179	–	–	–
Smoking	0.076	0.409	–	–	–
Age (years)	−0.279	0.002∗	−0.224	0.044	0.004∗
Height (cm)	−0.106	0.251	–	–	–
Body weight (kg)	0.002	0.981	–	–	–
Body mass index (kg/m2)	0.081	0.381	–	–	–
Systolic blood pressure (mmHg)	0.022	0.808	–	–	–
Diastolic blood pressure (mmHg)	0.123	0.180	–	–	–
Mean arterial pressure (mmHg)	0.084	0.360	–	–	–
Total cholesterol (mg/dL)	−0.239	0.009∗	–	–	–
Log-Triglyceride (mg/dL)	0.020	0.829	–	–	–
Log-HDL-C (mg/dL)	−0.009	0.923	–	–	–
LDL-C (mg/dL)	−0.304	0.001∗	–	–	–
Log-Glucose (mg/dL)	0.157	0.086	–	–	–
Log-BUN (mg/dL)	−0.160	0.082	–	–	–
Log-Creatinine (mg/dL)	−0.146	0.112	–	–	–
eGFR (mL/min)	0.200	0.028∗	–	–	–
Galectin-3 (ng/mL)	−0.512	<0.001∗	−0.387	0.255	<0.001∗

Data of triglyceride, HDL-C, fasting glucose, blood urea nitrogen, and creatinine showed skewed distribution and were log-transformed before analysis.

Data analysis was done using simple linear regression analyses or multivariable stepwise linear regression analysis (adapted factors were age, total cholesterol, LDL-C, eGFR, and galectin-3).

HDL-C, high-density lipoprotein cholesterol; LDL-C, low-density lipoprotein cholesterol; BUN, blood urea nitrogen; eGFR, estimated glomerular filtration rate.

∗p < 0.05 was considered statistically significant.

Table 3.

Multivariate logistic regression analysis for vascular reactivity dysfunction (intermediate vascular reactivity and poor vascular reactivity) or poor vascular reactivity in 120 hypertensive patients.

Model	Galectin-3 (per 1 ng/mL of increase) for vascular reactivity dysfunction		Galectin-3 (per 1 ng/mL of increase) for poor vascular reactivity
	OR (95 % CI)	p value	OR (95 % CI)	p value
Crude model	1.217 (1.101–1.344)	<0.001∗	1.225 (1.096–1.368)	<0.001∗
Model 1	1.214 (1.095–1.345)	<0.001∗	1.270 (1.112–1.450)	<0.001∗
Model 2	1.222 (1.092–1.366)	<0.001∗	1.312 (1.127–1.528)	<0.001∗
Model 3	1.222 (1.079–1.384)	0.002∗	1.373 (1.095–1.723)	0.006∗

Model 1: adjusted for age, sex, and body mass index. Model 2: Adjusted for Model 1 plus systolic blood pressure, diastolic blood pressure, fasting glucose, and eGFR. Model 3: Adjusted for Model 2 plus total cholesterol, triglyceride, HDL-C, LDL-C, anti-hypertensive drugs, and anti-lipid drugs.

HDL-C, high-density lipoprotein cholesterol; LDL-C, low-density lipoprotein cholesterol; eGFR, estimated glomerular filtration rate; OR, odds ratio; CI, confidence interval.

∗p < 0.05 was considered statistically significant.

3.3. Gal-3 is a predictor of cardiovascular outcomes in patients with hypertension

Because Gal-3 is significantly correlated with ED, we hypothesize that Gal-3 levels may be associated with CV outcomes in patients with hypertension. For this, we grouped the patients according to their Gal-3 levels and monitored the occurrence of CV events for 7 years. Of the 120 patients, 104 were under regular health surveillance in our hospital during the 7-year post-enrollment period. During the follow-up period, 39 patients had CV events (37.5 %). The CV events included CV death (n = 1, 1 %), acute coronary syndrome (n = 14, 13.5 %), heart failure (n = 3, 2.9 %), ischemic stroke (n = 3, 2.9 %) and hemorrhagic stroke (n = 3, 2.9 %). Patients who had CV events had significantly lower systolic BP (p = 0.049), diastolic BP (p = 0.007), mean arterial pressure (p = 0.010), eGFR (p = 0.010), and higher serum Gal-3 levels (p = 0.002) than those without CV events (Table 4). Cox regression analysis was performed to identify risk factors associated with CV events in patients with hypertension (Table 5). In the analysis, mean arterial pressure (hazard ratio [HR]: 0.964, 95 % CI, 0.939–0.989, p = 0.006), albumin (HR: 0.203, 95 % CI, 0.050–0.824, p = 0.026), eGFR (HR: 0.977, 95 % CI, 0.961–0.994, p = 0.008), and Gal-3 levels (HR: 2.391, 95 % CI, 1.261–4.531, p = 0.008) were associated with an increased risk of CV events in patients with hypertension. After adjustment for significant variables (mean arterial pressure, albumin, eGFR, and Gal-3) by multivariate Cox regression analysis, higher serum Gal-3 levels (adjusted HR: 1.072, 95 % CI, 1.013–1.134, p = 0.016), mean arterial pressure (adjusted HR: 0.974, 95 % CI, 0.947–0.996, p = 0.024), and albumin (adjusted HR: 0.161, 95 % CI, 0.030–0.849, p = 0.031) showed a significant association with CV events in patients with hypertension. Fig. 1 illustrates the survival probability based on serum Gal-3 levels, dividing patients into two groups according to a value greater or less than the mean value (12.87 ng/mL). Using a Kaplan–Meier curve, a higher serum Gal-3 level corresponded to a lower survival probability of CV events (p = 0.006), as determined by the log-rank test.

Table 4.

Clinical variables of the 104 hypertension patients with or without cardiovascular events.

Variables	All participants (n = 104)	Participants without CV events (n = 65)	Participants with CV events (n = 39)	p value
Age (years)	63.09 ± 7.35	62.18 ± 7.35	64.61 ± 7.18	0.103
Body mass index (kg/m2)	26.54 ± 3.72	26.16 ± 3.50	27.17 ± 4.03	0.183
Systolic blood pressure (mmHg)	138.50 ± 18.72	141.29 ± 17.95	133.85 ± 19.30	0.049∗
Diastolic blood pressure (mmHg)	81.08 ± 10.96	83.31 ± 9.67	77.36 ± 12.05	0.007∗
Mean arterial pressure (mmHg)	100.21 ± 12.40	102.64 ± 11.27	96.19 ± 13.28	0.010∗
Total cholesterol (mg/dL)	165.35 ± 39.79	162.43 ± 41.89	170.21 ± 36.01	0.337
Triglyceride (mg/dL)	124.00 (98.00–183.50)	122.00 (96.50–182.50)	126.00 (106.00–194.00)	0.699
HDL-C (mg/dL)	46.00 (40.00–55.00)	48.00 (39.00–55.00)	44.00 (41.00–56.00)	0.809
LDL-C (mg/dL)	94.05 ± 28.07	90.65 ± 25.70	99.72 ± 31.15	0.111
Fasting glucose (mg/dL)	109.00 (91.00–135.75)	109.00 (91.00–141.00)	105.00 (90.00–130.00)	0.689
Albumin (g/dL)	4.40 (4.20–4.50)	4.40 (4.20–4.50)	4.40 (4.20–4.50)	0.053
Blood urea nitrogen (mg/dL)	16.00 (14.00–19.00)	16.00 (13.50–18.00)	17.00 (14.00–22.00)	0.113
Creatinine (mg/dL)	1.00 (0.80–1.10)	1.00 (0.80–1.10)	1.00 (0.90–1.20)	0.089
eGFR (mL/min)	82.22 ± 22.31	86.55 ± 22.41	75.00 ± 20.43	0.010∗
Galectin-3 (ng/mL)	12.87 ± 4.87	11.74 ± 4.16	14.75 ± 5.41	0.002∗
Male, n (%)	86 (82.7)	55 (84.6)	31 (79.5)	0.503
Diabetes mellitus, n (%)	47 (45.2)	31 (47.7)	16 (41.0)	0.508
Coronary artery disease, n (%)	76 (73.1)	49 (75.4)	27 (69.2)	0.493
Smoking, n (%)	22 (21.2)	15 (23.1)	7 (17.9)	0.535
ACE inhibitor or ARB use, n (%)	71 (68.3)	46 (70.8)	25 (64.1)	0.479
β-blocker use, n (%)	66 (63.5)	38 (58.5)	28 (71.8)	0.172
CCB use, n (%)	42 (40.4)	29 (44.6)	13 (33.3)	0.256
Statin use, n (%)	81 (77.9)	50 (76.9)	31 (79.5)	0.760
Fibrate use, n (%)	8 (7.7)	7 (10.8)	1 (2.6)	0.128

Values for continuous variables were given as means ± standard deviation and compared by Student's t-test; variables not normally distributed (given as medians and interquartile range and compared by Mann-Whitney U test; values are presented as number (%), and analysis was performed using the chi-square test. CV, cardiovascular; HDL-C, high-density lipoprotein cholesterol; LDL-C, low-density lipoprotein cholesterol; eGFR, estimated glomerular filtration rate; ACE, angiotensin-converting enzyme; ARB, angiotensin-receptor blocker; CCB, calcium-channel blocker.

∗p < 0.05 was considered statistically significant.

Table 5.

Risk factors of cardiovascular events of the 104 hypertensive patients (multivariate Cox proportional-hazards regression analysis).

Variables	HR	95 % C.I.	p value	aHR	95 % C.I.	p value
Age (years)	1.032	0.986–1.080	0.177	–	–	–
Body mass index (kg/m2)	1.063	0.980–1.154	0.139	–	–	–
Female/Male	1.274	0.585–2.771	0.542	–	–	–
DM	0.751	0.397–1.421	0.379	–	–	–
CAD	0.755	0.382–1.490	0.417	–	–	–
Smoking	0.816	0.360–1.848	0.625	–	–	–
Mean arterial pressure (mmHg)	0.964	0.939–0.989	0.006∗	0.974	0.947–0.996	0.024∗
Galectin-3
<12.87 ng/mL	1			1
≥12.87 ng/mL	2.391	1.261–4.531	0.008∗	1.072	1.013–1.134	0.016∗
Total cholesterol (mg/dL)	1.004	0.997–1.011	0.291	–	–	–
Triglyceride (mg/dL)	1.002	0.999–1.004	0.130	–	–	–
HDL-C (mg/dL)	0.993	0.967–1.020	0.617	–	–	–
LDL-C (mg/dL)	1.008	0.998–1.017	0.104	–	–	–
Fasting glucose (mg/dL)	1.003	0.997–1.010	0.272	–	–	–
Albumin (g/dL)	0.203	0.050–0.824	0.026∗	0.161	0.030–0.849	0.031∗
eGFR (mL/min)	0.977	0.961–0.994	0.008∗	0.985	0.968–1.001	0.070
ACE inhibitor or ARB use	0.762	0.396–1.467	0.416	–	–	–
β-blocker use	1.578	0.785–3.172	0.200	–	–	–
CCB use	0.687	0.353–1.338	0.270	–	–	–
Statin use	1.039	0.477–2.261	0.923	–	–	–
Fibrate use	0.278	0.038–2.023	0.206	–	–	–

HR: hazard ratio; CI: confidence interval; aHR: adjusted hazard ratio; HDL-C, high-density lipoprotein cholesterol; LDL-C, low-density lipoprotein cholesterol; eGFR, estimated glomerular filtration rate; ACE, angiotensin-converting enzyme; ARB, angiotensin-receptor blocker; CCB, calcium-channel blocker.

∗p < 0.05 was considered statistically significant after Cox regression analysis. Multivariate Cox proportional-hazards regression analysis is adjusted for mean arterial pressure, albumin, eGFR, and galectin-3.

Fig. 1.

Kaplan-Meier analysis of galectin-3 (divided into two groups according to value more and less than means value) for cardiovascular events of hypertensive patients.

4. Discussion

In this study, serum Gal-3 level was negatively correlated with VRI and was an independent variable associated with ED in hypertensive patients. We found that higher serum Gal-3 levels at enrollment correlated with more CV events during the 7-year follow-up period in these patients. These data suggest that Gal-3, a variable well correlated with endothelial function, is a biomarker for predicting adverse CV events in patients with hypertension.

Gal-3 is predominantly distributed in the cytoplasm and nucleus [14] and regulates various intracellular signaling [28]. It is also released outside the cells and interacts with many membrane-bound receptors to carry out important extracellular functions [29]. In the circulation, Gal-3 can be handled by kidney and released into urine [30]. Is the Gal-3 protein in the patients’ serum an active modulator or just one of proteins released by the damaged cells? Oxidative stress can cause cell injury, which leads to the loss of cell membrane integrity and the release of cell contents into extracellular spaces [31,32]. Higher serum Gal-3 levels may represent a greater burden of cell damage, but not effective harmful mediators. However, Gal-3 may be secreted into the circulation in response to biochemical insults. A previous study suggested that alteration of the environment can modify the posttranslational modifications (PTMs) of Gal-3, which regulate the function of this protein and promote its secretion [33]. Although the mechanisms remain unclear, Gal-3 may be released into the circulation triggered by specific molecular signals, resulting in subsequent adverse CV events. If so, exploring the molecular signals that regulate Gal-3 secretion and identifying the specific clinical profiles in patients with high levels of Gal-3 will be helpful in the development of new strategies for promoting CV health.

Although Gal-3 is an emerging target in the field of CV diseases, its role in ED is not yet clear. Human data considering the correlation between Gal-3 levels and endothelial function are rare, and current available evidence about the role of Gal-3 in regulating endothelial homeostasis is scanty and controversial. We previously reported that serum Gal-3 levels were positively associated with the severity of ED in patients with chronic kidney disease [27]. These findings suggest that Gal-3 may be involved in modulating endothelial homeostasis in various disease settings. In the mouse model, ablation of Gal-3 protects endothelial cells from angiotensin II-induced stress [34] and obesity-related functional impairment [35]. Gal-3 was also proposed to aggravate ox-LDL-induced endothelial injury [36]. Conversely, Gal-3 deletion was found to aggravate ED in a DM mouse fed a high-fat diet [37]. In addition to directly modulating endothelial signaling, Gal-3 may participate in this regulation through inflammation. Studies have demonstrated that inflammation significantly contributes to the pathogenesis of ED [38]. Gal-3 is expressed by immune cells [39], regulates their activation [40], and promotes inflammatory responses [41]. These findings collectively suggest that Gal-3 plays multiple roles in the pathogenesis of ED, and further investigation is required.

Our study has some limitations. One concern is that the number of recruited patients in this study was relatively small and lack of healthy individuals. The source of patients has been confined to a single medical center. These findings may be a result of underrepresentation of diverse populations. Although the analysis of risk factors for CV outcomes (Table 5) showed that the presence of DM, CAD and the drug usage at enrollment are not independent risk factors for the subsequent CV events, the high prevalence of CAD and DM in our participants may still make it difficult to elaborate the role of Gal-3 in endothelial dysfunction and subsequent CV outcomes. And the lack of healthy subjects prevents us from examining the baseline serum Gal-3 levels in our general population. Another concern is that we only checked Gal-3 level once at enrollment but did not get its serial data during the follow-up period. We did not check other factors, such as oxidative stress or inflammation, which could also affect CV outcomes. The other concern is that although this study demonstrated the correlation of Gal-3, endothelial function, and CV outcomes in patients with hypertension, we did not have the data to prove its causal relationship. Therefore, we need more evidence from basic research to depict the role of Gal-3 in the pathogenesis of ED.

5. Conclusions

This study found that Gal-3 is significantly correlated with endothelial function in patients with hypertension. Higher Gal-3 levels are associated with worse endothelial function and unfavorable CV prognosis. These findings will provide us with a newer perspective of ED in individuals with hypertension.

CRediT authorship contribution statement

Hui-Sheng Wang: Writing – original draft, Methodology, Formal analysis, Data curation, Conceptualization. Bang-Gee Hsu: Writing – review & editing, Validation, Supervision, Methodology, Funding acquisition. Ji-Hung Wang: Validation, Resources, Investigation, Data curation. Chiu-Fen Yang: Writing – original draft, Methodology, Funding acquisition, Conceptualization.

Data availability

Data will be made available on request.

Funding sources

This study was supported by the Hualien Tzu Chi Hospital, Buddhist Tzu Chi Medical Foundation, Hualien, Taiwan (TCRD-112-060 and TCMF-A 106-01-08), and Academia Sinica Institutional Grand Challenge Project (Grant AS-GC-110-04).

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.