Abstract
Background
Circulating Galectin‐3 (Gal‐3) concentrations are associated with an increased incidence of heart failure, atrial fibrillation, chronic kidney disease, and mortality. Recent evidence suggests that Gal‐3 may also be an important modulator of cardiometabolic traits such as adiposity, insulin resistance, and hyperglycemia. We examined the associations of blood Gal‐3 concentrations and cardiometabolic disease traits in the Framingham Heart Study.
Methods and Results
In cross‐sectional analyses of 2946 Framingham Heart Study participants (mean age 59 years, 55% women), higher Gal‐3 concentrations were associated with higher body mass index, waist circumference, and triglycerides (P<0.0001 for all). Higher Gal‐3 was associated with greater odds of obesity (multivariable‐adjusted odds ratio 1.16 per 1‐SD increase in log–Gal‐3, 95% CI 1.06–1.28, P=0.002) and hypertension (odds ratio 1.18, 95% CI 1.07–1.29, P=0.0006). In prospective analyses, Gal‐3 was associated with incident metabolic syndrome (hazard ratio 1.22, 95% CI 1.10–1.36, P=0.0002) and diabetes (hazard ratio 1.21, 95% CI 1.04–1.41, P=0.02), in age‐ and sex‐adjusted, but not multivariable‐adjusted models.
Conclusions
In this large, community‐based sample, circulating Gal‐3 was associated with abdominal adiposity, dyslipidemia, and hypertension in cross‐sectional analyses, but Gal‐3 did not predict incident cardiometabolic disease after adjusting for cardiometabolic risk factors. Future investigations should focus on further elucidating mechanisms linking Gal‐3 with cardiometabolic disease and on assessing whether modulation of the Gal‐3 pathway might have positive cardiometabolic effects.
Keywords: epidemiology, obesity, prevention
Subject Categories: Epidemiology, Obesity
Introduction
Galectin‐3 (Gal‐3), a β‐galactoside–binding lectin, is expressed in several cell types, including epithelial and inflammatory cells, and is involved in important cellular processes such as adhesion, apoptosis, inflammation, and differentiation.1, 2 Gal‐3 also appears to modulate pathways central to numerous disease conditions, including fibrosis, inflammation, cancer progression, and metabolic disorders.2 Previous studies of Gal‐3 in cardiovascular disorders observed higher Gal‐3 concentrations to be associated with incident heart failure,3 atrial fibrillation,4 chronic kidney disease,5 atherosclerosis,6 and mortality7; however, the mechanisms linking Gal‐3 and cardiovascular disease are not well elucidated.
Cardiometabolic traits such as adiposity, diabetes, hypertension, and dyslipidemia are, in addition to smoking, the dominant modifiable risk factors for cardiovascular disease.8, 9 Recent experimental studies have demonstrated that Gal‐3 may be an important regulator of lipogenesis,1, 10 hyperglycemia,11 and obesity‐related inflammation.12, 13 Therefore, Gal‐3 may influence cardiovascular disease risk via its potential effects on cardiometabolic pathways. The association of Gal‐3 with cardiometabolic disease in humans is not well studied. We hypothesized that higher Gal‐3 would be associated with a greater burden of cardiometabolic traits in the community cross‐sectionally and with a greater incidence of cardiometabolic traits prospectively. We tested this hypothesis in a large community‐based epidemiological cohort in which participants have been well characterized with respect to cardiometabolic traits.
Methods
Study Sample
The enrollment and characteristics of the Framingham Heart Study (FHS) offspring cohort were reported previously.14 Briefly, 5124 children of the FHS original cohort and their spouses have been evaluated approximately every 4 years at the FHS research clinic since 1971. For the present investigation, 3532 participants attending the sixth offspring cohort examination cycle (1995–1998) were considered for analysis. Participants were excluded due to offsite visit (n=43), missing Gal‐3 measurement (n=41), extreme outlier Gal‐3 values (>5 log SDs [1.35 units] above or below the log‐transformed mean of 2.63 units; n=5), prevalent heart failure (n=38), prevalent major cardiovascular disease (n=144), stage IV chronic kidney disease (n=7), missing covariates (n=98), or nonfasting laboratory values (n=210), yielding a final sample of 2946 individuals. The Boston University Medical Center Institutional Review Board approved all study protocols, and all participants provided written informed consent.
Laboratory Measures
Fasting blood samples were collected, centrifuged, and immediately stored at −80°C during each examination cycle. Gal‐3 concentrations were measured in 2011 on EDTA plasma samples that were obtained during the sixth examination cycle (1995–1998) by using an enzyme‐linked immunosorbent assay (BG Medicine), which was previously demonstrated to display within‐run and total precision between 2.1% and 5.7% and 4.2% and 12.0%, respectively.15 The lower detection limit was 1.32 ng/mL and the upper quantification limit was 96.6 ng/mL. Serum triglyceride concentrations and high‐density lipoprotein (HDL) cholesterol concentrations were measured by using automated enzymatic assays after precipitation of apolipoprotein B–containing lipoproteins with dextran‐sulfate magnesium. Serum glucose was measured at each examination cycle by using a hexokinase reagent. Insulin concentrations were measured during the fifth examination cycle (1991–1995) with a radioimmunoassay (Coat‐A‐Count Insulin; Diagnostic Products). The Homeostasis Model Assessment of Insulin Sesistance (HOMA‐IR) was calculated by the formula: HOMA‐IR=fasting plasma insulin [microunits per milliliter]×(fasting plasma glucose [millimoles per liter])/22.5.16 Insulin resistance was defined as a value above the top quartile of HOMA‐IR from the whole FHS Offspring sample free of diabetes [75th percentile of HOMA‐IR=2.74 (mg/dL)×(IU/mL)].17 C‐reactive protein (CRP) was measured at the sixth examination cycle with a Dade Behring nephelometer.
Clinical Assessment and Outcome Definitions
A cardiovascular‐focused physical examination, anthropometry, and phlebotomy were performed at each examination cycle at the FHS clinic. Systolic and diastolic blood pressure (BP) measurements were recorded as the mean of 2 physician readings. Hypertension was defined as systolic BP ≥140 mm Hg, diastolic BP ≥90 mm Hg, or use of antihypertension medications. Body mass index (BMI) was calculated as weight in kilograms divided by the square of height in meters, and obesity was defined as BMI ≥30 kg/m2. Diabetes was defined as fasting serum glucose ≥126 mg/dL or receiving drug treatment for elevated glucose. Metabolic syndrome was defined as the presence of any 3 of the following 5 criteria: (1) waist circumference ≥40 inches in men or ≥35 inches in women; (2) triglycerides ≥150 mg/dL; (3) HDL cholesterol <40 mg/dL in men or <50 mg/dL in women; (4) systolic BP ≥130 mm Hg or diastolic BP ≥85 mm Hg or drug treatment for hypertension; and (5) fasting glucose ≥100 mg/dL or drug treatment for elevated glucose.18 A physical activity index composite score was calculated as the sum of the reported time spent performing specific levels of physical activity during a 24‐hour period, weighted by the estimated oxygen consumption for each activity, as described previously.19 For prospective analyses, incident metabolic syndrome, diabetes, obesity, and hypertension were defined using subsequent clinical assessments during the seventh (1998–2001) and eighth (2005–2008) examination cycles.
Statistical Analysis
Gal‐3 was natural log–transformed for all analyses due to a right‐skewed distribution. Baseline characteristics were displayed according to sex‐specific Gal‐3 quartiles. Multivariable linear regression models were used to relate Gal‐3 with the following cardiometabolic traits: BMI, waist circumference, triglycerides (log‐transformed), HDL cholesterol, fasting glucose, HOMA‐IR, and systolic and diastolic BPs. Because HOMA‐IR was ascertained at a previous examination cycle, the sample size of eligible participants with available measurements was lower (n=2691). For BP analyses, values were corrected for concomitant antihypertensive treatment by adding 15 mm Hg to the systolic BP and 10 mm Hg to the diastolic BP.20 Models were first adjusted for age and sex, and then additionally adjusted for systolic BP, antihypertensive treatment status, BMI, diabetes status, current smoking status, physical activity, and triglycerides (log‐transformed). We excluded the dependent variable or its close relation for each model (ie, BMI and waist circumference analyses were not adjusted for BMI, fasting glucose and HOMA‐IR were not adjusted for diabetes status, systolic and diastolic BP were not adjusted for systolic BP or antihypertensive treatment status).
Multivariable logistic regression models were used to evaluate the associations of Gal‐3 with the following cardiometabolic binary traits: metabolic syndrome, diabetes, obesity, hypertension, and insulin resistance. For evaluation of the association of Gal‐3 with metabolic syndrome, we included an intermediate model, adjusted for clinical risk factors that are not part of the definition of metabolic syndrome itself: age, sex, physical activity, and smoking.
To evaluate the prospective association of Gal‐3 concentrations with incident metabolic syndrome, diabetes, obesity, and hypertension during the 2 follow‐up examinations, we used discrete‐time Cox proportional hazards regression models. Participants with prevalent disease were excluded for each analysis, yielding different sample sizes for each outcome. We adjusted each model as follows: (1) age and sex and (2) age, sex, smoking status, physical activity, systolic BP, antihypertensive treatment status, fasting blood glucose, HDL cholesterol, triglycerides (log‐transformed), and BMI. In addition, for metabolic syndrome analyses, an intermediate model including age, sex, smoking status, and physical activity but excluding variables that are part of the definition of metabolic syndrome was constructed.
Penalized splines were created to test for nonlinearity in the multivariable‐adjusted associations of Gal‐3 with all outcomes (cross‐sectional and prospective). In a sensitivity analysis, models for traits related to dysglycemia and insulin resistance (ie, fasting blood glucose, HOMA‐IR, diabetes, and insulin resistance) were repeated after excluding participants with prevalent diabetes. In exploratory analyses, we evaluated the relations of Gal‐3 with inflammation as assessed by CRP in the 2852 individuals with both biomarkers measured. CRP was natural log‐transformed for these analyses. The correlation between Gal‐3 and CRP concentrations was assessed by using the Spearman correlation coefficient. We added CRP to multivariable models to see whether inflammation (CRP) partially explained the relations of Gal‐3 with cross‐sectional cardiometabolic traits.
For the primary analyses, we tested 5 distinct traits that comprise metabolic syndrome: waist circumference, BP, HDL cholesterol, triglycerides, and hyperglycemia. Therefore, we used a Bonferroni‐corrected P‐value threshold of 0.01 to denote statistical significance and considered P<0.05 to be suggestive. For prospective analyses, a P‐value of 0.05 was considered statistically significant. All analyses were performed using SAS Software version 9.3 (SAS Institute).
Results
A total of 2946 individuals were included, with mean age of 59 years and 55% women. Baseline clinical characteristics by sex‐specific Gal‐3 quartiles are shown in Table 1. The mean age and proportion of individuals with diabetes, obesity, hypertension, and insulin resistance increased across Gal‐3 quartile.
Table 1.
Characteristics of the Study Sample by Gal‐3 Quartile
| Characteristic | Quartile 1 (n=730) | Quartile 2 (n=746) | Quartile 3 (n=729) | Quartile 4 (n=741) |
|---|---|---|---|---|
| Age, y | 55±9 | 58±9 | 60±9 | 63±9 |
| Women, n (%) | 404 (55) | 411 (55) | 399 (55) | 403 (54) |
| BMI, kg/m2 | 26.9±4.8 | 27.7±5.2 | 28.2±5.2 | 28.7±5.2 |
| Waist circumference, cm | 37±5 | 38±5 | 38±5 | 39±5 |
| Obesity, n (%) | 156 (21) | 199 (27) | 209 (29) | 251 (34) |
| Heart rate, beats per minute | 63±10 | 63±10 | 64±10 | 65±10 |
| Systolic blood pressure, mm Hg | 124±18 | 127±18 | 130±19 | 132±20 |
| Diastolic blood pressure, mm Hg | 75±9 | 76±9 | 76±9 | 76±10 |
| Antihypertensive medication use, n (%) | 107 (15) | 161 (22) | 201 (28) | 305 (41) |
| Hypertension, n (%) | 201 (28) | 265 (36) | 309 (42) | 405 (55) |
| Smoking, n (%) | 102 (14) | 116 (16) | 119 (16) | 98 (13) |
| Diabetes, n (%) | 50 (7) | 56 (8) | 59 (8) | 93 (13) |
| Fasting blood glucose, mg/dL | 100±24 | 102±26 | 103±25 | 105±28 |
| Insulin resistance, n (%)a | 139 (19) | 187 (25) | 211 (29) | 245 (33) |
| HOMA‐IR, mg×IU/dL×mLa | 2.6±10.4 | 2.4±3.1 | 2.5±3.3 | 3.3±5.5 |
| Triglycerides, mg/dL | 118±68 | 127±78 | 143±93 | 158±96 |
| HDL cholesterol, mg/dL | 54±17 | 52±17 | 51±16 | 49±15 |
| Metabolic syndrome, n (%) | 221 (30) | 280 (38) | 317 (44) | 386 (52) |
| C‐reactive protein, mg/Lb | 1.5 (0.7, 3.5) | 1.7 (0.8, 3.6) | 2.5 (1.0, 4.8) | 3.0 (1.3, 6.8) |
| Gal‐3, ng/mLc | 10.2 (9.3, 10.9) | 12.5 (12.0, 13.2) | 14.7 (14.1, 15.6) | 18.0 (16.9, 20.0) |
Values are mean±SD or median (25th percentile, 75th percentile) for continuous variables, and n (%) for categorical variables. BMI indicates body mass index; Gal‐3, galectin‐3; HDL, high‐density lipoprotein; HOMA‐IR, Homeostatic Model Assessment‐Insulin Resistance.
HOMA‐IR measurements were available in 2691 participants.
C‐reactive protein concentrations were available in 2852 participants.
Sex‐specific Gal‐3 values by quartile, in ng/mL, for men: Q1: 3.9 to 10.9; Q2: 11.0 to 12.8; Q3: 12.9 to 15.1; Q4: 15.2 to 54.4; and for women: Q1: 5.0 to 11.9; Q2: 12.0 to 14.2; Q3: 14.3 to 16.7; Q4: 16.8 to 52.1.
Cross‐Sectional Associations of Gal‐3 With Continuous Cardiometabolic Traits
The associations of Gal‐3 concentrations with continuous cardiometabolic traits are displayed in Table 2. In age‐ and sex‐adjusted analyses, Gal‐3 concentrations were positively associated with BMI, waist circumference, serum triglyceride concentrations, HOMA‐IR, and systolic and diastolic BP and inversely related to HDL cholesterol (P<0.0001 for all). The associations with BMI, waist circumference, and triglycerides persisted after multivariable adjustment (P<0.0001 for all). In sensitivity analyses, we examined the relations of fasting glucose and HOMA‐IR after excluding those with diabetes, and the results were similar (data not shown).
Table 2.
Cross‐Sectional Associations of Gal‐3 With Continuous Cardiometabolic Traits
| Cardiometabolic Trait | Age and Sex Adjusted (β Estimate±SE) | P Value | Multivariable Adjusted (β Estimate±SE) | P Value |
|---|---|---|---|---|
| Body mass index, kg/m2 | 0.79±0.10 | <0.0001 | 0.47±0.10 | <0.0001 |
| Waist circumference, cm | 2.18±0.26 | <0.0001 | 1.32±0.25 | <0.0001 |
| Triglycerides (log‐transformed) | 0.10±0.01 | <0.0001 | 0.07±0.01 | <0.0001 |
| HDL cholesterol, mg/dL | −2.16±0.29 | <0.0001 | −0.49±0.26 | 0.06 |
| Fasting glucose, mg/dL | 0.96±0.50 | 0.05 | −1.09±0.49 | 0.03 |
| HOMA‐IR (log‐transformed) | 0.12±0.02 | <0.0001 | 0.04±0.02 | 0.03 |
| Systolic blood pressure, mm Hg | 1.57±0.38 | <0.0001 | 0.51±0.38 | 0.18 |
| Diastolic blood pressure, mm Hg | 1.03±0.21 | <0.0001 | 0.38±0.21 | 0.07 |
Regression coefficients represent the mean change associated with a 1‐SD higher log‐transformed Gal‐3 (0.27 unit), which is equivalent to an increase in Gal‐3 from 13.87 ng/mL (the sample mean) to 18.17 ng/mL. Multivariable model adjusts for age, sex, systolic blood pressure (except for systolic blood pressure and diastolic blood pressure), antihypertensive treatment status (except for systolic blood pressure and diastolic blood pressure), body mass index (except for body mass index and waist circumference), diabetes status (except for fasting glucose and HOMA‐IR), current smoking, physical activity, and log‐transformed triglycerides (except for triglycerides). Gal‐3 indicates galectin‐3; HDL, high‐density lipoprotein; HOMA‐IR, Homeostatic Model Assessment‐Insulin Resistance.
Cross‐Sectional Associations of Gal‐3 With Binary Cardiometabolic Traits
Each 1‐SD increase in log‐transformed Gal‐3 was associated with higher odds of having metabolic syndrome in age‐ and sex‐adjusted analyses (odds ratio [OR] 1.29, 95% confidence interval 1.19–1.40, P<0.0001), Table 3. When additionally adjusted for smoking and physical activity, Gal‐3 remained associated with metabolic syndrome (OR 1.28, 95% CI 1.17–1.39, P<0.0001); however, the full multivariable‐adjusted model (including covariates used to define metabolic syndrome) was not statistically significant (OR 0.91, 95% 0.80–1.02, P=0.11).
Table 3.
Cross‐Sectional Associations of Gal‐3 With Binary Cardiometabolic Traits
| Cardiometabolic Trait | Age and Sex Adjusted | Multivariable Adjusted | ||
|---|---|---|---|---|
| OR (95% CI) | P Value | OR (95% CI) | P Value | |
| Metabolic syndrome | 1.29 (1.19 to 1.40) | <0.0001 | 0.91 (0.80 to 1.02) | 0.11 |
| Obesity | 1.30 (1.19 to 1.42) | <0.0001 | 1.16 (1.06 to 1.28) | 0.002 |
| Diabetes | 1.22 (1.06 to 1.40) | 0.004 | 1.01 (0.87 to 1.18) | 0.87 |
| Insulin resistance | 1.25 (1.14 to 1.37) | <0.0001 | 1.02 (0.91 to 1.13) | 0.77 |
| Hypertension | 1.29 (1.19 to 1.41) | <0.0001 | 1.18 (1.07 to 1.29) | 0.0006 |
OR represent the change associated with a 1‐SD higher log‐transformed Gal‐3. Multivariable model adjusts for age, sex, systolic blood pressure (except for hypertension), antihypertensive treatment status (except for hypertension), body mass index (except for obesity), diabetes status (except for diabetes), current smoking, physical activity, and log‐transformed triglycerides. Gal‐3 indicates galectin‐3; OR, odds ratio.
In age‐ and sex‐ adjusted models, we observed statistically significant direct relations for Gal‐3 concentrations with diabetes (OR 1.22, 95% CI 1.06–1.40, P=0.004), obesity (OR 1.30, 95% CI 1.19–1.42, P<0.0001), hypertension (OR 1.29, 95% CI 1.19–1.41, P<0.0001), and insulin resistance (OR 1.25, 95% CI 1.14–1.37, P<0.0001), Table 3. The associations with obesity (OR 1.16, 95% CI 1.06–1.28, P=0.002) and hypertension (OR 1.18, 95% CI 1.07–1.29, P=0.0006) remained statistically significant after further adjustment for cardiometabolic risk factors, whereas the associations with diabetes and insulin resistance were attenuated. Penalized splines did not reveal nonlinearity for any of the significant associations of Gal‐3 and cardiometabolic traits.
Associations of Gal‐3 With Incident Cardiometabolic Disease
Of the 1612 participants free of metabolic syndrome at the baseline (sixth) examination, 280 developed metabolic syndrome by the seventh examination cycle (1579 [98%] participants attended the seventh examination, mean time interval from baseline examination 2.9 years), and 227 developed metabolic syndrome by the eighth examination cycle (1460 [91%] participants attended the eighth examination, mean time interval from baseline examination 9.5 years). In age‐ and sex‐adjusted Cox proportional hazards models, each 1‐SD increase in log‐transformed Gal‐3 was associated with higher risk of incident metabolic syndrome (hazard ratio [HR] 1.22, 95% CI 1.10–1.36, P=0.0002; Table 4). The association persisted after adjustment for age, sex, smoking, and physical activity (HR 1.16, 95% CI 1.04–1.29, P=0.007) but was no longer significant after further multivariable adjustment for cardiometabolic risk factors (HR 1.02, 95% CI 0.90–1.16, P=0.74).
Table 4.
Associations of Gal‐3 With Incident Cardiometabolic Disease
| Cardiometabolic Trait | No. Events/n at Risk | Age and Sex Adjusted | Multivariable Adjusted | ||
|---|---|---|---|---|---|
| HR (95% CI) | P Value | HR (95% CI) | P Value | ||
| Metabolic syndrome | 507/1612 | 1.22 (1.10 to 1.36) | 0.0002 | 1.02 (0.90 to 1.16) | 0.74 |
| Obesity | 247/1982 | 1.13 (0.99 to 1.28) | 0.07 | 0.96 (0.82 to 1.13) | 0.65 |
| Diabetes | 187/2513 | 1.21 (1.04 to 1.41) | 0.02 | 1.08 (0.88 to 1.32) | 0.49 |
| Hypertension | 734/1694 | 1.03 (0.94 to 1.14) | 0.51 | 1.00 (0.90 to 1.12) | 0.96 |
HR represents the hazard ratio for each 1‐SD higher log‐transformed Gal‐3. Multivariable model adjusts for age, sex, systolic blood pressure, antihypertensive treatment status, body mass index, diabetes status, current smoking, physical activity, and log‐transformed triglycerides. Gal‐3 indicates galectin‐3.
Among 2513 individuals free of diabetes at the baseline examination, diabetes developed in 72 participants by the seventh examination cycle (2460 [98%] participants attended the seventh examination, mean time interval from baseline examination 2.9 years) and in 115 participants by the eighth examination cycle (2203 [88%] participants attended the eighth examination, mean time interval from baseline examination 9.5 years). In age‐ and sex‐adjusted analyses, each 1‐SD increase in Gal‐3 was associated with increased incidence of diabetes (HR 1.21, 95% CI 1.04–1.41, P=0.02). This association was not statistically significant after multivariable adjustment for cardiometabolic risk factors. Gal‐3 concentrations did not predict incident obesity or hypertension.
Relations of Gal‐3 and CRP
Gal‐3 concentrations modestly correlated with CRP in our sample (Spearman correlation coefficient 0.25, P<0.0001). When the multivariable models assessing the associations of Gal‐3 with cross‐sectional cardiometabolic traits were additionally adjusted for CRP, we observed partial attenuation of the effect estimates. The association of Gal‐3 with obesity was no longer statistically significant after further adjustment for CRP (OR 1.08, 95% CI 0.98–1.19, P=0.13). The associations of Gal‐3 and the continuous variables of BMI (P=0.01) and waist circumference (P=0.0021) were partially attenuated but remained statistically significant. The relations of Gal‐3 with triglycerides (P<0.0001) and hypertension (OR 1.15, 95% CI 1.05–1.27, P=0.0037) were also partially attenuated by further adjustment for CRP.
Discussion
In our community‐based investigation, circulating concentrations of Gal‐3 were positively associated with several cardiometabolic traits, including abdominal adiposity, dyslipidemia, and hypertension, in cross‐sectional analyses. Prospective associations of Gal‐3 with incident diabetes and metabolic syndrome were no longer significant after full adjustment for cardiometabolic risk factors.
The observed relations of Gal‐3 with cardiometabolic traits may partially explain the previous reports of associations between Gal‐3 concentrations and cardiovascular diseases including heart failure and atherosclerosis.3, 4, 5, 6, 7, 21, 22, 23 Although not a primary focus, several of these previous studies reported associations of Gal‐3 with cardiometabolic traits. For example, in the FHS, Gal‐3 concentrations were associated hypertension treatment status and BMI.3 Similarly in the PREVEND (Prevention of REnal and Vascular ENd‐stage Disease) cohort, Gal‐3 concentrations were related to BP, BMI, HDL cholesterol, and triglycerides. However, these associations were not evaluated in multivariable models adjusting for other cardiometabolic risk factors.7
The role of Gal‐3 in cardiometabolic disease has also been evaluated in experimental models of adiposity, insulin resistance, diabetes, and inflammation. In mouse models, Gal‐3 promotes adipocyte differentiation through activation of peroxisome proliferator‐activated receptor‐γ, a nuclear receptor involved in adipogenesis,1 while Gal‐3–deficient mice display accelerated rates of diet‐induced obesity, hyperglycemia, insulin resistance, and systemic inflammation.11, 13 Further, Gal‐3 expression in adipocytes can be induced by proinflammatory signals such as circulating free fatty acids and interleukin‐6.12 In a study of ≈100 people, Gal‐3 concentrations were directly correlated with diabetes status.24 However, among diabetics, galectin‐3 concentrations were paradoxically found to be inversely related to hemoglobin A1c. Whether Gal‐3 is a mediator of inflammation and cardiometabolic disease remains unclear. Our exploratory analyses show modest correlation with inflammation, and further studies are needed to explore underlying pathways.
Previous studies have evaluated the associations of novel biomarkers with cardiometabolic traits. In cross‐sectional studies, several adipokines (most notably adiponectin and leptin) have been correlated with obesity and hypertension,25, 26, 27, 28 but prospective reports are conflicting.29 Similarly, although inflammatory markers have been linked with diabetes pathogenesis,30 a previous report including 2600 FHS participants demonstrated that a panel of 12 inflammatory biomarkers was associated with incident diabetes in age‐ and sex‐adjusted, but not fully adjusted, models.31 The present investigation, therefore, suggests that Gal‐3 may play a similar role as these previous biomarkers in cardiometabolic disease; Gal‐3 is related to cardiometabolic traits, but its use as a predictor of future cardiometabolic risk is likely limited. We have previously demonstrated an association of Gal‐3 and incident heart failure and all‐cause mortality.3 While Gal‐3 remained significantly associated with these clinical outcomes after multivariable adjustment, performance metrics evaluating reclassification showed a limited role for Gal‐3 in risk prediction.
Several limitations of our investigation merit consideration. First, although experimental data have demonstrated potential direct effects of Gal‐3 on obesity and metabolic disease, causal inferences cannot be drawn from the results of this observational study. Second, Gal‐3 concentrations were measured at 1 time point (the baseline examination), whereas changes in this value over time might provide further information. Third, all cross‐sectional traits were measured contemporaneously with Gal‐3 with the exception of HOMA‐IR, which was assessed at the preceding examination cycle, resulting in a reduced sample size. This might introduce bias if the individuals with missing HOMA‐IR values differed from the broader sample. The noncontemporaneous measures might also allow for a regression dilution bias, which would likely underestimate a true association. Fourth, our sample was composed primarily of Caucasian individuals of European descent; the generalizability to other population is therefore unknown.
In conclusion, Gal‐3 concentrations are associated with cross‐sectional measures of abdominal adiposity, dyslipidemia, and hypertension in a large community‐based sample. However, Gal‐3 did not predict incident cardiometabolic traits after adjusting for traditional clinical risk factors. Therefore, our findings do not support a role for Gal‐3 as a clinical biomarker of cardiometabolic risk in the community. Future investigations should focus on elucidating the mechanisms linking Gal‐3 to cardiometabolic traits and on examining whether modulation of the Gal‐3 pathway might improve cardiometabolic outcomes.
Sources of Funding
This work was partially supported by the National Heart, Lung, and Blood Institute's Framingham Heart Study (contracts N01‐HC‐25195 and HHSN268201500001I) and K23‐HL116780 (Dr Ho). Dr Ho is supported by a Boston University School of Medicine, Department of Medicine Career Investment award (Boston, MA). Dr Nayor received support from National Institutes of Health grant T32‐HL007604.
Disclosures
Galectin‐3 assays were provided by BG Medicine (Waltham, MA). BG Medicine had no access to study data and no input into data analyses, interpretation, or preparation of the manuscript. The authors have no other relevant disclosures to report.
(J Am Heart Assoc. 2016;5:e002347 doi: 10.1161/JAHA.115.002347)
References
- 1. Baek J‐H, Kim S‐J, Kang HG, Lee H‐W, Kim J‐H, Hwang K‐A, Song J, Chun K‐H. Galectin‐3 activates PPARγ and supports white adipose tissue formation and high‐fat diet‐induced obesity. Endocrinology. 2015;156:147–156. [DOI] [PubMed] [Google Scholar]
- 2. Pugliese G, Iacobini C, Pesce CM, Menini S. Galectin‐3: an emerging all‐out player in metabolic disorders and their complications. Glycobiology. 2015;25:136–150. [DOI] [PubMed] [Google Scholar]
- 3. Ho JE, Liu C, Lyass A, Courchesne P, Pencina MJ, Vasan RS, Larson MG, Levy D. Galectin‐3, a marker of cardiac fibrosis, predicts incident heart failure in the community. J Am Coll Cardiol. 2012;60:1249–1256. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Ho JE, Yin X, Levy D, Vasan RS, Magnani JW, Ellinor PT, McManus DD, Lubitz SA, Larson MG, Benjamin EJ. Galectin 3 and incident atrial fibrillation in the community. Am Heart J. 2014;167:729–734.e1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. O'Seaghdha CM, Hwang S‐J, Ho JE, Vasan RS, Levy D, Fox CS. Elevated galectin‐3 precedes the development of CKD. J Am Soc Nephrol. 2013;24:1470–1477. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Madrigal‐Matute J, Lindholt JS, Fernandez‐Garcia CE, Benito‐Martin A, Burillo E, Zalba G, Beloqui O, Llamas‐Granda P, Ortiz A, Egido J, Blanco‐Colio LM, Martin‐Ventura JL. Galectin‐3, a biomarker linking oxidative stress and inflammation with the clinical outcomes of patients with atherothrombosis. J Am Heart Assoc. 2014;3:e000785 doi: 10.1161/JAHA.114.000785. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. de Boer RA, van Veldhuisen DJ, Gansevoort RT, Muller Kobold AC, van Gilst WH, Hillege HL, Bakker SJL, van der Harst P. The fibrosis marker galectin‐3 and outcome in the general population. J Intern Med. 2012;272:55–64. [DOI] [PubMed] [Google Scholar]
- 8. Mozaffarian D, Benjamin EJ, Go AS, Arnett DK, Blaha MJ, Cushman M, de Ferranti S, Després J‐P, Fullerton HJ, Howard VJ, Huffman MD, Judd SE, Kissela BM, Lackland DT, Lichtman JH, Lisabeth LD, Liu S, Mackey RH, Matchar DB, McGuire DK, Mohler ER, Moy CS, Muntner P, Mussolino ME, Nasir K, Neumar RW, Nichol G, Palaniappan L, Pandey DK, Reeves MJ, Rodriguez CJ, Sorlie PD, Stein J, Towfighi A, Turan TN, Virani SS, Willey JZ, Woo D, Yeh RW, Turner MB; American Heart Association Statistics Committee and Stroke Statistics Subcommittee . Heart disease and stroke statistics‐2015 update: a report from the American Heart Association. Circulation. 2015;131:e29–e322. [DOI] [PubMed] [Google Scholar]
- 9. Singh GM, Danaei G, Farzadfar F, Stevens GA, Woodward M, Wormser D, Kaptoge S, Whitlock G, Qiao Q, Lewington S, Di Angelantonio E, vander Hoorn S, Lawes CMM, Ali MK, Mozaffarian D, Ezzati M; Global Burden of Metabolic Risk Factors of Chronic Diseases Collaborating Group; Asia‐Pacific Cohort Studies Collaboration (APCSC), Diabetes Epidemiology: Collaborative analysis of Diagnostic criteria in Europe (DECODE), Emerging Risk Factor Collaboration (ERFC), Prospective Studies Collaboration (PSC) . The age‐specific quantitative effects of metabolic risk factors on cardiovascular diseases and diabetes: a pooled analysis. PLoS One. 2013;8:e65174. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Kiwaki K, Novak CM, Hsu DK, Liu F‐T, Levine JA. Galectin‐3 stimulates preadipocyte proliferation and is up‐regulated in growing adipose tissue. Obesity (Silver Spring). 2007;15:32–39. [DOI] [PubMed] [Google Scholar]
- 11. Pang J, Rhodes DH, Pini M, Akasheh RT, Castellanos KJ, Cabay RJ, Cooper D, Perretti M, Fantuzzi G. Increased adiposity, dysregulated glucose metabolism and systemic inflammation in galectin‐3 KO mice. PLoS One. 2013;8:e57915. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Krautbauer S, Eisinger K, Hader Y, Buechler C. Free fatty acids and IL‐6 induce adipocyte galectin‐3 which is increased in white and brown adipose tissues of obese mice. Cytokine. 2014;69:263–271. [DOI] [PubMed] [Google Scholar]
- 13. Pejnovic NN, Pantic JM, Jovanovic IP, Radosavljevic GD, Milovanovic MZ, Nikolic IG, Zdravkovic NS, Djukic AL, Arsenijevic NN, Lukic ML. Galectin‐3 deficiency accelerates high‐fat diet‐induced obesity and amplifies inflammation in adipose tissue and pancreatic islets. Diabetes. 2013;62:1932–1944. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Kannel WB, Feinleib M, McNamara PM, Garrison RJ, Castelli WP. An investigation of coronary heart disease in families. The Framingham offspring study. Am J Epidemiol. 1979;110:281–290. [DOI] [PubMed] [Google Scholar]
- 15. Christenson RH, Duh S‐H, Wu AHB, Smith A, Abel G, deFilippi CR, Wang S, Adourian A, Adiletto C, Gardiner P. Multi‐center determination of galectin‐3 assay performance characteristics: anatomy of a novel assay for use in heart failure. Clin Biochem. 2010;43:683–690. [DOI] [PubMed] [Google Scholar]
- 16. Matthews DR, Hosker JP, Rudenski AS, Naylor BA, Treacher DF, Turner RC. Homeostasis model assessment: insulin resistance and beta‐cell function from fasting plasma glucose and insulin concentrations in man. Diabetologia. 1985;28:412–419. [DOI] [PubMed] [Google Scholar]
- 17. Preis SR, Massaro JM, Robins SJ, Hoffmann U, Vasan RS, Irlbeck T, Meigs JB, Sutherland P, D'Agostino RB, O'Donnell CJ, Fox CS. Abdominal subcutaneous and visceral adipose tissue and insulin resistance in the Framingham Heart Study. Obesity (Silver Spring). 2010;18:2191–2198. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Grundy SM, Cleeman JI, Daniels SR, Donato KA, Eckel RH, Franklin BA, Gordon DJ, Krauss RM, Savage PJ, Smith SC, Spertus JA, Costa F. Diagnosis and management of the metabolic syndrome: an American Heart Association/National Heart, Lung, and Blood Institute Scientific Statement. Circulation. 2005;112:2735–2752. [DOI] [PubMed] [Google Scholar]
- 19. Kannel WB, Sorlie P. Some health benefits of physical activity. The Framingham Study. Arch Intern Med. 1979;139:857–861. [PubMed] [Google Scholar]
- 20. Tobin MD, Sheehan NA, Scurrah KJ, Burton PR. Adjusting for treatment effects in studies of quantitative traits: antihypertensive therapy and systolic blood pressure. Stat Med. 2005;24:2911–2935. [DOI] [PubMed] [Google Scholar]
- 21. Grandin EW, Jarolim P, Murphy SA, Ritterova L, Cannon CP, Braunwald E, Morrow DA. Galectin‐3 and the development of heart failure after acute coronary syndrome: pilot experience from PROVE IT‐TIMI 22. Clin Chem. 2012;58:267–273. [DOI] [PubMed] [Google Scholar]
- 22. Maiolino G, Rossitto G, Pedon L, Cesari M, Frigo AC, Azzolini M, Plebani M, Rossi GP. Galectin‐3 predicts long‐term cardiovascular death in high‐risk patients with coronary artery disease. Arterioscler Thromb Vasc Biol. 2015;35:725–732. [DOI] [PubMed] [Google Scholar]
- 23. Iacobini C, Menini S, Ricci C, Scipioni A, Sansoni V, Cordone S, Taurino M, Serino M, Marano G, Federici M, Pricci F, Pugliese G. Accelerated lipid‐induced atherogenesis in galectin‐3‐deficient mice: role of lipoxidation via receptor‐mediated mechanisms. Arterioscler Thromb Vasc Biol. 2009;29:831–836. [DOI] [PubMed] [Google Scholar]
- 24. Weigert J, Neumeier M, Wanninger J, Bauer S, Farkas S, Scherer MN, Schnitzbauer A, Schäffler A, Aslanidis C, Schölmerich J, Buechler C. Serum galectin‐3 is elevated in obesity and negatively correlates with glycosylated hemoglobin in type 2 diabetes. J Clin Endocrinol Metab. 2010;95:1404–1411. [DOI] [PubMed] [Google Scholar]
- 25. Stepien M, Stepien A, Banach M, Wlazel RN, Paradowski M, Rizzo M, Toth PP, Rysz J. New obesity indices and adipokines in normotensive patients and patients with hypertension: comparative pilot analysis. Angiology. 2014;65:333–342. [DOI] [PubMed] [Google Scholar]
- 26. Stepien M, Stepien A, Wlazel RN, Paradowski M, Rizzo M, Banach M, Rysz J. Predictors of insulin resistance in patients with obesity: a pilot study. Angiology. 2014;65:22–30. [DOI] [PubMed] [Google Scholar]
- 27. Stepien M, Wlazeł RN, Paradowski M, Banach M, Rysz M, Misztal M, Rysz J. Serum concentrations of adiponectin, leptin, resistin, ghrelin and insulin and their association with obesity indices in obese normo‐ and hypertensive patients—pilot study. Arch Med Sci. 2012;8:431–436. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28. Ingelsson E, Larson MG, Yin X, Wang TJ, Meigs JB, Lipinska I, Benjamin EJ, Keaney JF, Vasan RS. Circulating ghrelin, leptin, and soluble leptin receptor concentrations and cardiometabolic risk factors in a community‐based sample. J Clin Endocrinol Metab. 2008;93:3149–3157. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. Vinitha R, Ram J, Snehalatha C, Nanditha A, Shetty AS, Arun R, Godsland IF, Johnston DG, Ramachandran A. Adiponectin, leptin, interleukin‐6 and HbA1c in the prediction of incident type 2 diabetes: a nested case–control study in Asian Indian men with impaired glucose tolerance. Diabetes Res Clin Pract. 2015;109:340–346. [DOI] [PubMed] [Google Scholar]
- 30. Pejnovic NN, Pantic JM, Jovanovic IP, Radosavljevic GD, Djukic AL, Arsenijevic NN, Lukic ML. Galectin‐3 is a regulator of metaflammation in adipose tissue and pancreatic islets. Adipocyte. 2013;2:266–271. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31. Dallmeier D, Larson MG, Wang N, Fontes JD, Benjamin EJ, Fox CS. Addition of inflammatory biomarkers did not improve diabetes prediction in the community: the Framingham Heart Study. J Am Heart Assoc. 2012;1:e000869 doi: 10.1161/JAHA.112.000869. [DOI] [PMC free article] [PubMed] [Google Scholar]