Abstract
Background:
Hypertriglyceridemia is an independent risk factor for major adverse cardiovascular events, though the mechanisms linking triglycerides and platelet function with thrombosis, remain elusive. The aim of this study was to assess the association between platelet function and triglyceride levels.
Methods:
We included participants from the Framingham Heart Study Third Generation cohort, OMNI, and New Offspring Spouse cohort who attended the third examination cycle (2016–2019). Eligible participants were categorized into four triglyceride subgroups.
Results:
The study comprised a total of 1897 (55.53%) participants with normal TG levels; 883 (25.85%) participants with high-normal TGs; 378 (11.07%) with borderline high TGs; and 258 (7.55%) participants with hypertriglyceridemia. After adjusting for age, sex, alcohol consumption, aspirin, statin and P2Y12 inhibitors, the levels of ADP-induced platelet aggregation were inversely associated with total cholesterol levels (P<0.0001). Platelet disaggregation was associated with low-density lipoprotein and high-density lipoprotein cholesterol levels (P<0.0001). Lastly, in a shear-stress chamber assay mimicking arterial flow velocities, TG levels in the normal-high group were associated with increased levels of collagen-dependent thrombogenicity (β=24.16, SE=6.65, P<0.0001).
Conclusion:
Triglyceride levels are associated with altered platelet activation and aggregation. Furthermore, increased platelet-driven thrombogenicity is directly associated with triglyceride levels after adjusting for medications and other covariates.
Keywords: platelets, triglycerides, granules, thrombus
Introduction
Cardiovascular disease (CVD) accounts for more than 16% of all-cause mortality in the United States (US) (1). Notably, data from the National Health and Examination Survey highlights the persistent ethnic disparities in the prevalence of dyslipidemia and cardiovascular health-related outcomes in the US (2). Dyslipidemia remains a major independent risk factor for coronary artery disease (CAD) (3), and in particular, hypertriglyceridemia (HTG) is a strong independent predictor for the occurrence of a major adverse cardiovascular event (4,5). In the past three decades, the prevalence of dyslipidemia in the US has gradually increased, with over 30% of the adult population having elevated fasting serum triglyceride levels (>150mg/dL)(6,7).
Lipoproteins are a target for initiating lipid-lowering therapy, however, their use in CVD risk-stratification remains contentious (8,9). Interestingly, triglyceride (TG) levels modify the classification of cardiovascular risk regardless of high-density lipoprotein cholesterol (HDL-c) and low-density lipoprotein cholesterol (LDL-c) levels (10). However, the precise mechanisms linking HTG with CAD remain unclear. Activated platelets seem to be one plausible link between dyslipidemia and atherosclerosis (10,11). In addition, plasma triglyceride levels may predict platelet responsiveness to aspirin following an ischemic stroke (12). Despite the previously reported inconsistent and pleiotropic effects of LDL-c and HDL-c lipoproteins on circulating platelets (13,14), the binding of lipoproteins to the surface of circulating platelets alters platelet alpha and dense granule release (15).
To date, there is no consensus on the monitoring of platelet function using light transmission aggregometry (LTA) in HTG. However, LTA remains the gold standard for assessing platelet function in most clinical settings. Since serum lipid levels may affect LTA (16), a comprehensive platelet phenotyping and functional analysis approach is required. This may enable the elucidation of the functional changes that occur in peripheral blood platelets in patients with HTG. Therefore, we assessed the association between platelet reactivity and triglyceride levels in the large community-based Framingham Heart Study (FHS). We provide the first comprehensive platelet phenotyping using several platelet assays and functional analysis of platelet integrin activation, degranulation and thrombogenicity in HTG.
Method
Study sample description
A detailed description of the FHS study design and participant selection criteria has been described elsewhere (17). In this study we included participants from the Third Generation cohort, the New Offspring Spouse (NOS) and OMNI cohorts who attended the third examination cycle (2016–2019), and also had corresponding lipid profiles measurements and platelet function assayed during the selected exam. Briefly, the Third Generation cohort and the older, smaller NOS cohort comprised of 98.4% white participants of European ancestry [17,18]. We excluded participants who had missing lipid measurements (n=1) or those with lipidemic samples (n=6) and participants (n=12) who had extremely elevated triglyceride levels (≥600 mg/dL). The study was approved by the Boston University Institutional Review Board, and written informed consent was obtained from each participant.
Classification of triglyceride subgroups
Eligible participants were categorized into four subgroups based on the Adult Treatment Panel III (ATP III) TG categories (6,18). This comprised of; (I) normal and low-normal TG (reference group) (TG<100 mg/dL); (II) high-normal TG (100–149 mg/dL); (III) borderline-high TG (150–199 mg/dL); (IV) high TG (>200 mg/dL).
Lipid and biochemical measurements
The lipid profiles were determined using a Cobas 501 instrument (Roche Diagnostics GmbH, Mannheim, Germany) run on EDTA plasma samples (Tyco Monoject 15% EDTA (K3). C-reactive protein (CRP) was measured on serum from glass serum vacutainers(Becton Dickinson, Franklin Lakes, NJ). Specimens were centrifuged for 22 minutes at 2500×g at 4° C to separate serum and plasma within 90 minutes of draw. All measurements were done with Roche kits: cholesterol (Catalog #03039773), HDL (Catalog #04399803), TGs (Catalog #20767107), HbA1c (Catalog #05336163), glucose (Catalog #04404483), and CRP (Catalog #04628918).
Platelet function measurements
Platelet function is characterized by three distinct phases: integrin activation, degranulation, aggregation, and adhesion. To provide a comprehensive platelet function phenotyping, platelet integrin activation (αIIbβ3) and degranulation (αgranule) were measured using flow cytometry, with PAC-1 and P-selectin/CD62P, respectively after stimulation with ADP. Platelet reactivity to ADP, AA and collagen-induced aggregation was measured using LTA, whole blood Multiplate® impedance aggregometry (MP) and the Optimul 96-well plate assay Optimul (19). We also determined thrombogenicity using the microchip-based Total Thrombus Formation Analysis System (T-TAS) (Zacros, Tokyo, Japan), where platelets form thrombus in collagen-coated microchambers resulting in pressure gradient changes. Detailed descriptions of the sample collection and preparation procedures (including PRP preparation) and analytical methods for platelet functional assays used have been described elsewhere (20). All samples were brought into the laboratory immediately after collection. No pneumatic tube systems were used, as these are known to affect platelet function by multiplate impedance aggregometry (21,22). Platelet assays were performed within 3 hours of sampling.
Platelet integrin activation and degranulation
To assess ADP-induced alpha and dense granule release and αIIbβ3 integrin activation, we measured the changes in P-selectin/CD62P and PAC-1 expression following stimulation with 20μM of ADP. Briefly, citrated whole blood samples (3.2% sodium citrate) and PRP samples were stimulated with 20μM of ADP and incubated for 15 minutes at 37 ° C. The samples were then stained using a titrated volume of an antibody cocktail containing anti-PAC-1 FITC(AK-4), anti-CD 61 PerCP-Cy5.5(Clone VI-PL2; 1:20), and anti-CD62P APC (Clone AK-4;1:5) (all from BD Biosciences, San Jose, CA, USA). The samples were then incubated for 15 minutes at 37 °C and subsequently fixed using the BD Cytofix ™ fixation buffer (BD Biosciences, San Jose, CA, USA). A primary data acquisition stop gate was set at 10,000 CD61+ events, and data was acquired at a low flow rate (to minimize coincidental CD61+ events) on the BD Accuri C6 flow cytometer (BD Biosciences, San Jose, CA, USA). The 8-peak rainbow calibration beads (Spherotech, Lake Forest, IL, USA) were used in instrument calibration and to monitor the instrument’s optical path and laminar flow. Single-stained whole blood samples were used to determine the levels of spectral overlap and to perform color compensation. Analytic gates were drawn based on SSC/CD61+ events along with the corresponding marker (PAC-1, CD62P) and random samples were repeatedly examined throughout the 3-year period to fine tune the gates. All ADP stimulation experiments were normalized within sample from the same blood tube against vehicle/normal saline controls that were similarly stained and fixed, to derive a % activation change. For quality control of the acquired flow cytometry data, we used flowAI (23) which is an automatic and interactive anomaly discerning tool to perform quality control of our acquired data and to exclude anomalies due to abrupt changes in flow rate, signal acquisition instability, and outlies based on dynamic range of each fluorescence channel (23).
Measurement of platelet aggregation
Light transmission and Multiplate impendence aggregometry
Platelet aggregation was assessed in PRP using LTA and in whole blood using MP impendence aggregometry. Briefly, for LTA we used varying concentrations of ADP (0.95μM, 1.82 μM, 5.71μM), and a single concentration of AA (500μg/mL), collagen (190 μg/mL), ristocetin (1.5mg/mL), Epinephrine (0.1mmol/L), TRAP-6 amine (0.67mM). LTA was used to determine platelet aggregation in response to ADP, AA, collagen receptor-mediated responses. In this assay, platelet-rich plasma (PRP) was prepared from citrated (3.2% sodium citrate) whole blood samples. Briefly, a two-step differential centrifugation method was followed, and samples were initially centrifuged for 10 minutes at 200×g to yield P-PRP with subsequent centrifugation at 1500×g for 15 minutes to yield platelet poor plasma (PPP) (16). Platelet LTA was performed using an eight-channel PAP-8E platelet aggregometer (Bio/Data Corporation, Horsham. PA, USA). Briefly, 225μl of PRP was pipetted into each cuvette with a micro stir bar and incubated at 37 °C for 2 minutes. Full aggregation (100%) was set by normalization of each well to a cuvette of isogenic 225μl PPP plus 25μl normal saline. Lastly, 25μl of agonist stock was added to the respective response cuvettes, and platelet aggregation was measured at 300 rpm over 6 minutes. In LTA, the primary aggregation, percentage disaggregation and primary slope and AUC over time for each agonist was determined and reported as a measure of platelet aggregation (AUC*min).
Optimul 96-well plate assay
Optimul aggregometry was performed as previously described (19,20) and platelet responses to varying concentrations of AA (0.03mM-1.0mM), ADP (0.005–40μM), collagen (0.01μg/ml-40 μg/ml), epinephrine (0.0004–10μmol/L), ristocetin (0.14–4 mg/mL), TRAP-6 (0.03–40μmol/L) and U46619 (0.005μM-40 μM) were measured. In Optimul aggregometry, the mean AUC over time for a range of concentrations for each agonist was reported as a measure of platelet aggregation.
Measurement of total thrombogenicity with the T-TAS
To assess platelet-collagen mediated thrombogenicity, we used T-TAS (Zacros, Tokyo, Japan). The microchip-based flow chamber system allows for the measurement of thrombus formation under physiological arterial shear rates ranging between 1000–3000 per reciprocal second (s−1). In this study, we measured total thrombogenicity using the platelet chip (PL-chip), which contains 25 capillary channels ~50 uM in width that are coated with type-1 collagen (24). Briefly, 320μl of whole blood was perfused through the PL-chip at a shear rate of 1500s−1, and the area underneath the pressure wave-form at 10 minutes (PL-AUC10) was used as a measure of collagen-mediated thrombogenicity.
Statistical analysis
In our analysis, we firstly considered lipid profiles as continuous variables and explored the associations between lipid levels and platelet reactivity. Since the distribution of TG levels is markedly skewed(18), we further analyzed platelet reactivity using well-defined clinical triglyceride categories, as recommended (18). Age, sex, aspirin use, and statin use adjusted means of lipid profiles were compared across the TG subgroups using a linear mixed effects model (LME) that treats each pedigree as a cluster with lme4 (https://cran.r-project.org/web/packages/lme4/) and emmeans (https://cran.r-project.org/web/packages/emmeans/) R packages, and the global p-values based on a three degrees of freedom Wald Chi-Square statistic were computed to compare the mean lipid profiles among the TG groups (Table 2). We tested the association between platelet function and the clinical TG subgroups using the LME adjusting for age, sex, aspirin, and statin use (Table 3). To assess the relationship between lipid profiles and platelet function, we first applied inverse normal transformation to platelet function and then tested the association between transformed trait with lipid profiles using LME adjusting for age, sex, BMI, fasting status, alcohol consumption, the use of aspirin, P2Y12 inhibitors and statins (Figure 1, Supplementary Table 1–3). All statistical analyses were performed using the R language and environment (25). To correct for multiple comparisons, a Bonferroni-correction threshold of 0.0011 was used.
Table 2.
Baseline characteristics of participants in triglyceride subgroups (n=3416).
| Normal TG levels (n=2780) | Elevated TG levels (n=636) | P-Value | |||
|---|---|---|---|---|---|
| Referent group I | Group II Normal- High | Group III | Group IV | ||
| (TG levels<100) | (TG 100–150 mg/dL) | Borderline High (TG >150–199 mg/dL) | High (TG>200 mg/dL) | ||
| N | 1897 | 883 | 378 | 258 | |
| Age | 53.83(9.42) | 55.54(0.32) | 55.08(0.45) | 54.14(0.54) | 0.0034 |
| Sex | <0.0001 | ||||
| Male, n(%) | 724(38.20) | 476(53.91) | 204(53.96) | 173(67.05) | |
| Female, n(%) | 1173(61.80) | 407(46.09) | 174(46.03) | 85(32.95) | |
| Male: Female ratio | 1.62 | 1.17 | 1.17 | 2.03 | |
| BMI | 26.79(5.39) | 29.95(0.19) | 32.12(0.29) | 31.36(0.35) | <0.0001 |
| Diabetes | 47(2.48) | 68(7.70) | 47(12.43) | 39(15.11) | <0.0001 |
| Alcohol consumption (drinks/week) | 5.02(0.21) | 4.92(0.26) | 4.06(0.374) | 5.58(0.44) | 0.0353 |
| Fasting, n (%) | 1840(97.00) | 764(96.46) | 369(97.62) | 246(95.35) | 0.4138 |
| Non fasting, n (%) | 57(3.00) | 28(3.53) | 9(2.38) | 12(4.65) | |
| Time Fasting (minutes) | 757.90(120.27) | 766.31(4.28) | 770.15(6.32) | 754.55(8.85) | 0.1482 |
| BP measurements a | |||||
| SBP | 117.55(1.18) | 120.52(1.22) | 121.76(1.22) | 124.77(1.37) | <0.0001 |
| DBP | 73.07(0.74) | 75.30(0.82) | 75.58(0.82) | 78.23(0.86) | <0.0001 |
| Glucose metabolism a | |||||
| Plasma glucose, mmol | 104.87(1.82) | 108.87(1.87) | 114.60(2.01) | 119.33(2.01) | <0.0001 |
| HbA1c, % | 5.67(0.0.6) | 5.78(0.06) | 6.00(0.07) | 6.06(0.07) | <0.0001 |
| Lipid Profiles a | |||||
| TC, mg/dL | 160.58(2.95) | 174.48(3.03) | 179.23(3.25) | 189.44(3.40) | <0.0001 |
| LDL-c, mg/dL | 85.43(2.68) | 100.54(2.75) | 102.53(2.94) | 96.59(3.14) | <0.0001 |
| HDL-c, mg/dL | 61.11(1.37) | 49.80(1.41) | 42.91(1.51) | 39.36(1.58) | <0.0001 |
| TG, mg/dL | 72.02(2.36) | 123.04(2.43) | 171.54(2.61) | 276.23(2.74) | <0.0001 |
| Medication, No.(%) | |||||
| Statins | 309(16.29) | 238(26.95) | 129(34.13) | 92(35.66) | <0.0001 |
| Aspirin | 357(18.82) | 219(24.80) | 81(21.43) | 73(28.29) | 0.0001 |
| P2Y12 inhibitors | 10(0.53) | 9(1.08) | 6(1.59) | 7(2.71) | 0.0121 |
Age, sex, aspirin, statin, P2Y12 inhibitor use- adjusted means. BP: Blood pressure; SBP: Systolic blood pressure; DBP: Diastolic blood pressure, TC: Total Cholesterol, LDL-c: Low-density lipoprotein cholesterol, HDL-c: High-density lipoprotein cholesterol, TG: Triglycerides.
Table 3.
Age, sex, aspirin and statin use adjusted multivariable regression analysis of platelet reactivity in triglyceride subgroups.
| Platelet trait | Assay | N | Group II: Normal-High TG 100–150 mg/dL) |
Group III: Borderline-High TG 150–199 mg/dL) |
Group IV: High TG >200 mg/dL) |
|||
|---|---|---|---|---|---|---|---|---|
| β(SE) | P-value | β(SE) | P-value | β(SE) | P-value | |||
|
| ||||||||
| ADPlow (AUC) | LTA | 3125 | −9.27 (4.56) | 0.0422 | −14.53(6.31) | 0.0214 | −25.00(7.4) | 0.0021 |
| ADPmid (AUC) | LTA | 3322 | −12.86(5.02) | 0.0105 | −25.63(6.93) | <0.001 | −31.71(8.2) | <0.0001 |
| MPadp (AUC) | MP | 3399 | 4.51(0.80) | <0.0001 | 1.95(1.1) | 0.0761 | 3.361(1.3) | 0.0096 |
| MPaspi (AUC) | MP | 3260 | 3.67 (0.91) | <0.0001 | 3.48(1.3) | 0.0578 | 3.95(1.5) | 0.0078 |
| %Disaggregation | LTA | 3322 | 1.42 (0.32) | <0.0001 | 2.05(0.4) | <0.0001 | 2.26(0.5) | <0.0001 |
| Ristocetin (AUC) | MP | 3361 | 5.08(1.13) | <0.0001 | 3.43(1.6) | 0.0281 | 6.92(1.8) | <0.0001 |
| MPtrap (AUC) | MP | 3400 | 7.16(0.88) | <0.0001 | 6.9(1.2) | <0.0001 | 11.24(1.4) | <0.0001 |
| TTAS (AUC) | TTAS | 1120 | 24.16(6.55) | <0.0001 | 19.0(8.9) | 0.0326 | −4.607(10.3) | 0.655 |
Significant values (p<0.005) shown in boldface.
AUC: Area under curve; TTAS: Total thrombus formation.
Fig. 1.
Forest plot of platelet-lipid trait associations from linear mixed effects models after adjusting for age, sex, BMI, fasting status, alcohol consumption, and the use of aspirin, P2Y12 inhibitors and statins. AA=arachidonic acid, ADP=adenosine diphosphate, Aspi=Multiplate Aspirin (AA) test, AUC=area-under-the-curve, LTA=light transmission aggregometry, MP=Multiplate impedance aggregometry in whole blood, PLT=platelet, PRP=platelet rich plasma, TRAP-6=thrombin receptor activator pepide 6, WB=whole blood.
Results
The cross-sectional analysis included 3416 participants who attended the third examination cycle (2016–2019). The population sample consisted of participants who were self-reported as Asian (n=69), African (n=55), Hispanic (n=28), Native Hawaiian or Pacific Islander (n=1), White (n=3072) and the race of some participants was unknown or mixed (Supplementary Table 1). The baseline characteristics of the sample population, which comprised of 1577 males and 1839 females are described in Table 1. There were no significant differences in the age of the included participants (p>0.05). Notably, relatively more males were on aspirin (29.96%) and fewer were on statins (21.37%) when compared to females. (Table 1)
Table 1.
Baseline characteristics of included participants (n= 3416) from the Third generation cohort, New Offspring Spouse (NOS) and OMNI cohorts, third exam cycle (2016–2019).
| Male | Female | p-value | |
|---|---|---|---|
|
|
|||
| (n=1577) | (n=1839) | ||
|
| |||
| Age (years) | 54.64(0.23) | 54.25(0.22) | 0.2395 |
| BMI | 29.52(0.13) | 27.71(0.15) | <0.0001 |
| SBP | 122.81(0.33) | 116.83(0.34) | <0.0001 |
| DBP | 78.11(0.21) | 73.66(0.19) | <0.0001 |
| Glucose metabolism | |||
| Plasma glucose, mmol/L | 105.23(0.63) | 95.98(0.41) | <0.0001 |
| HbA1c, % | 5.50(0.02) | 5.34(0.01) | <0.0001 |
| Lipid profiles, mg/dL | |||
| TC | 182.42(0.90) | 195.07(0.84) | <0.0001 |
| LDL-c | 107.55(0.81) | 107.87(0.75) | 0.6789 |
| HDL-c | 50.46(0.38) | 67.41(0.45) | <0.0001 |
| TG | 123.32(1.82) | 99.09(1.25) | <0.0001 |
| Statins, No. (%) | 337(21.37) | 431(23.45) | <0.0001 |
| Aspirin, No. (%) | 441(29.96) | 289(15.71) | <0.0001 |
BMI: Body max index; SBP: Systolic blood pressure; DBP: Diastolic blood pressure; HbA1c: Glycosylated haemoglobin; TC: Total cholesterol; LDL-c: Low density lipoprotein-c; HDL-c: High-density lipoprotein-c; TG: Triglycerides. Data are presented as mean and standard error (SE).
Total cholesterol (TC) levels are inversely associated with ADP-induced platelet aggregation and associated with TRAP-6 induced platelet aggregation.
There were significant associations between total cholesterol levels (TC) and the flow cytometry-based, PRP platelet count (β= 0.558, SE=0.104), and whole blood estimated platelet count (β= 0.282, SE=0.070, P<0.0001) (Figure 1, Supplementary Table 2). However, there were no associations between TC and the levels of ADP-induced platelet integrin activation (PAC-1) and degranulation (CD62P), P>0.05 (Figure 1, Supplementary Table 2). The levels of TC were inversely associated with ADP-induced platelet aggregation (P<0.0001) (Figure 1, Supplementary Table 2&3) and directly associated with TRAP-6 induced platelet aggregation (β=0.0502, SE=0.011, P<0.0001) (Figure 1, Supplementary Table 3).
High-density lipoprotein cholesterol (HDL-c) levels are inversely associated with TRAP-6, AA and ristocetin-induced platelet aggregation.
HDL-c levels were inversely associated with platelet reactivity to TRAP-6 (β=−0.1571, SE=0.024, P<0.0001), AA (β=−0.1349, SE=0.025, P<0.0001) and ristocetin (β=−0.1087, SE=0.031, P=0.0004) in MP impedance aggregometry (Figure 1, Supplementary Table 1). Interestingly, HDL-c levels were not associated with any of the agonist-induced percentage primary aggregation platelet traits (P>0.05).
Low-density lipoprotein cholesterol (LDL-c) levels are inversely associated with ADP, AA-induced platelet aggregation.
LDL-c and TG levels were inversely associated with decreased levels of primary platelet aggregation in response to ADP (P<0.001) (Figure 1, Supplementary Table 4). Notably, only LDL-c levels were associated with decreased AA-induced primary platelet aggregation (β=−0.013, SE=0.004, P<0.004). Lastly, HDL-c and LDL-c levels were directly associated with increased platelet disaggregation following ADP-induced platelet activation (P<0.0001) and TG were only associated with increased disaggregation following activation with 1.82 μM of ADP (P<0.0001) (Supplementary Table 2).
Triglyceride levels are associated with decreased ADP-induced aggregation and increased ristocetin and TRAP-6-induced aggregation
In LTA Triglyceride levels were inversely associated with platelet aggregation in response to ADP (P<0.0001). Interestingly, in MP impedance aggregometry, triglyceride levels were directly associated with platelet aggregation in response to Ristocetin (β=0.028, SE=0.008, P<0.0003) and TRAP-6 (β=0.050, SE=0.006, P<0.0001) (Figure 1, Supplementary Table 1).
A total of 3416 participants were then categorized into TG subgroups. These included a reference group (n=1897); high-normal (n=883); a borderline-high (n=356), and a high TG group (n=253). In addition, 16.29% of participants in the reference group were on statins and 18.82% were on aspirin. The C-reactive protein levels were high (>3mg/L) in the participants with borderline-high (150–199 mg/dL) or high TG levels (>200 mg/dL), when compared to the reference group (TG<100 mg/dL) (p<0.0001 and p=0.0001, respectively).
Platelet activation and aggregation in hypertriglyceridemia
In the multivariable regression model adjusted for age, sex, aspirin, P2Y12-inhibitors and statin use, there were no significant differences in the levels of peripheral blood leukocytes, platelet-leukocyte aggregates, ADP-induced integrin activation (PAC-1), and platelet degranulation (CD62P) between the TG groups (data not shown). Notably, in the high TG group, platelet aggregation was significantly decreased in response to a submaximal concentration of ADP (1.82 μM) (β=−31.71, SE=8.2, p< 0.0001) (Table 3). Interestingly, in impedance aggregometry, platelet reactivity increased in the high-normal group in response to ADP (3.19 μM) (β=4.51, SE=0.80, p< 0.0001) and AA (β=3.67, SE=0.91, p< 0.0001). In addition, platelet reactivity to ristocetin and TRAP-6 was increased in the high TG group (p<0.0001) (Table 3).
Platelet glycoprotein VI-mediated platelet aggregation and thrombogenicity
To assess platelet glycoprotein VI-mediated responses, we measured platelet aggregation and thrombogenicity levels in response to collagen. Notably, the levels of thrombogenicity in T-TAS shear-stress chambers were associated with increasing TG levels in participants with high-normal TG levels (β=24.159, SE=6.54, p<0.001). However, there were no associations between TG levels and collagen-mediated thrombogenicity in the high TG groups (Table 3).
Discussion
The aim of the study was to assess the association between platelet function and TG levels in the FHS Generation Three, OMNI and NOS cohorts. In a previous FHS study comprised of the Offspring cohort, TG levels modified the CVD-risk regardless of HDL and LDL levels (10). In fact, borderline high TG levels (above 150 mg/dl) were independently associated with a 10–20% increase in CVD (10). Prior studies on platelet function in dyslipidemia are limited to LTA, which may be affected by chylomicrons (CMs) (16,26). While other studies have also reported on the inverse association between HDL-c (27) and platelet activation, the potential modifying effect of TG levels has largely been neglected. In this study, 25.9% (n=883) of adult participants had normal but slightly elevated TG levels (100–150mg/dL), while 18.6% (n=636) of the participants were classified as HTG, defined as (TG levels>150mg/dL). The current study presents the largest population-based study evaluating the association between platelet function and TG levels, via several different assay approaches.
In our study, the levels of platelet P-selectin/CD62P expression upon activation, were comparable across all TG groups. Our null findings on the association between TG levels with platelet integrin activation and degranulation are in line with two previous studies that found no association between platelet integrin activation and TG levels (11,28). Similar findings have been reported in a cross-sectional study of patients with HTG (11). Notably, a negative association between platelet activation and TG levels has been reported (29). A plausible explanation may involve mechanisms leading to the rapid disassociation of membrane P-selectin into its soluble form following activation (30). Platelet reactivity in response to ADP and AA, was inversely associated with TC levels, LDL-c and HDL-c. Similar findings were reported in a previous study by Kameda et al.(31), where the levels of HDL -c were inversely associated with platelet reactivity to ADP and collagen. In this study we also report an association between decreased platelet reactivity to TRAP-6 and HDL-c levels. Notably, HDL-c levels are known to be associated with decreased platelet aggregation (27).
LTA remains a gold standard for assessing in vivo platelet aggregation, however, there is no clear consensus on the use and interpretation of aggregometry in the context of HTG (16). Although several guidelines have been published on the technical pre-analytical considerations of LTA, the influence of sample turbidity (particularly on P-PPP normalization steps) on LTA remains a challenging technical consideration for samples collected from patients with hypertriglyceridemia. Notably in impedance aggregometry, the lipemia of samples does not seem to be a critical factor that influences the accuracy of the test results (32).
In our study, participants with high TG levels (>200mg/dL) had lower levels of ADP-induced platelet aggregation by LTA when compared to the individuals with normal TG levels (<100mg/dL). Notably, the levels of platelet disaggregation in response to a low concentration of ADP (0.98μMol/L) were increased in participants with hypertriglyceridemia, this may explain the inverse association between the platelet aggregation and TG levels in these individuals. Platelet aggregation in response to endogenous agonists such as ADP is a complex process that involves active counter-responses, a reversible component of aggregation (disaggregation)[31]. Platelet disaggregation is a measure of the stability of platelet aggregates, and increased levels of disaggregation indicate low stability of platelet aggregates(34).
In contrast, we observed a trend toward elevated levels of platelet aggregation in participants with high TG levels (>200 mg/dL) and a significant increase in platelet aggregation in the Normal High group (100–150 mg/dL) by impedance aggregometry. The differences in the concentrations of ADP, sample preparation methods, and cellular milieu in the test samples may explain these incongruent results between LTA and impedance aggregometry. In addition, platelet aggregation in response to ADP may be influenced by other cellular components in whole blood, including lipids, red blood cells, and leukocytes (35). The findings based on LTA may be partially explained by the well-described interference of chylomicrons (CMs) with LTA (16). In fact, CMs are triglyceride-rich lipoproteins (36), and in the context of HTG, their interference with LTA may be substantial. Unlike LTA, whole blood impedance aggregometry can be used to assess platelet reactivity in lipemic and hemolyzed samples(32,37). Impedance aggregometry also measures platelet reactivity in the presence of white blood cells and red blood cells closely mimicking the physiological platelet microenvironment in peripheral blood circulation.
Finally, we evaluated collagen-mediated platelet responses in HTG using LTA based methods and the relatively novel T-TAS system. In our study, collagen-receptor mediated platelet aggregation by LTA was not associated with TG levels. However, elevated TG levels are associated with enhanced collagen mediated thrombus formation in high-normal TG levels. The T-TAS assay provides an artificial environment that may more closely align with in vivo biology in regard to whole blood platelet-arterial plaque formation than the stir bar, dilute agonist and PPP/PRP assay approach of LTA. Since the T-TAS assay relies on changes in pressure gradients as platelet-driven thrombus form on collagen, it is not susceptible to potential confounding of light measurements due to CMs or other factors in HTG that may affect light transmission or absorbance assays. These findings provide insight and potential platelet mediated mechanisms that may explain the recently reported association between TG levels and CVD-risk at levels previously considered within the normal range (<150mg/dL) (38).
The current study has several limitations. First, our study was comprised predominantly of those with European ancestry, which limits extrapolation of these findings to different non-European cohorts. Ethnic differences in TG levels exist, for example African American adults are known to have lower TG compared to other ethnic groups (7,39). Since lipid-lowering drugs such as statins and fibrates may inhibit platelet function, future studies providing a comprehensive subgroup analysis of platelet integrin activation, degranulation, and thrombogenicity are necessary under a variety of sample conditions.
Conclusion
Our findings suggest that TG levels may be associated with altered platelet aggregation at TG levels that are considered as normal (100–150mg/dL). Furthermore, platelet-driven thrombogenicity is directly associated with TG levels (at a cut-off value of 100–150mg/dL) and this further supports the previously described independent association between TG and increased thrombotic risk.
Supplementary Material
Acknowledgements
The authors would like to thank the participants and staff at the Framingham Heart Study.
Funding
This research was primarily supported by a special Population Sciences funding award to Dr. Andrew Johnson from the NHBLI Intramural Research program. The Framingham Heart Study (FHS) acknowledges the support of Contracts NO1-HC-25195, HHSN268201500001I and 75N92019D00031 from the National Heart, Lung and Blood Institute and grants HL107385, HL126136, HL93328, HL 142983, HL143227 and HL 131532 for this research.
The views expressed in this manuscript are those of the authors and do not necessarily represent the views of the National Heart, Lung, and Blood Institute; the National Institutes of Health; or the US Department of Health and Human Services.
Footnotes
Conflicts of interest
The authors have no competing interests or conflicts to declare.
Declaration of interests
The authors declare no conflict of interest. The funders had no role in the design of the study, collection, analyses or interpretation of the data, writing of the manuscript or in the decision to publish the reported results.
References
- 1.Virani SS, Alonso A, Aparicio HJ, et al. Heart disease and stroke statistics—2021 update: a report from the American Heart Association. Circulation. 2021;143(8):e254–e743. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Pool LR, Ning H, Lloyd-Jones DM, et al. Trends in racial/ethnic disparities in cardiovascular health among US adults from 1999--2012. J. Am. Heart Assoc. 2017;6(9):e006027. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Wong ND. Epidemiological studies of CHD and the evolution of preventive cardiology. Nat. Rev. Cardiol. 2014;11(5):276. [DOI] [PubMed] [Google Scholar]
- 4.Nordestgaard BG, Benn M, Schnohr P, et al. Nonfasting triglycerides and risk of myocardial infarction, ischemic heart disease, and death in men and women. Jama. 2007;298(3):299–308. [DOI] [PubMed] [Google Scholar]
- 5.Bansal S, Buring JE, Rifai N, et al. Fasting compared with nonfasting triglycerides and risk of cardiovascular events in women. Jama. 2007;298(3):309–316. [DOI] [PubMed] [Google Scholar]
- 6.National Cholesterol Education Program (US). Expert, of High Blood Cholesterol in Adults T. Third report of the National Cholesterol Education Program (NCEP) Expert Panel on detection, evaluation, and treatment of high blood cholesterol in adults (Adult Treatment Panel III). The Program; 2002. [Google Scholar]
- 7.Ford ES, Giles WH, Dietz WH. Prevalence of the Metabolic Syndrome Among US Adults. 2002;287(3):14–17. [DOI] [PubMed] [Google Scholar]
- 8.Puri R, Nissen SE, Shao M, et al. Non-HDL Cholesterol and Triglycerides: Implications for Coronary Atheroma Progression and Clinical Events. Arterioscler. Thromb. Vasc. Biol. 2016;36(11):2220–2228. [DOI] [PubMed] [Google Scholar]
- 9.Singh K, Chandra A, Sperry T, et al. Associations Between High-Density Lipoprotein Particles and Ischemic Events by Vascular Domain, Sex, and Ethnicity: A Pooled Cohort Analysis. Circulation. 2020;142(7):657–669. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Bartlett J, Predazzi IM, Williams SM, et al. Is isolated low high-density lipoprotein cholesterol a cardiovascular disease risk factor? Circ. Cardiovasc. Qual. Outcomes. 2016;9(3):206–212. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.de Man FH, Nieuwland R, van der Laarse A, et al. Activated platelets in patients with severe hypertriglyceridemia: effects of triglyceride-lowering therapy. Atherosclerosis. 2000;152(2):407–414. [DOI] [PubMed] [Google Scholar]
- 12.Karepov V, Tolpina G, Kuliczkowski W, et al. Plasma triglycerides as predictors of platelet responsiveness to aspirin in patients after first ischemic stroke. Cerebrovasc. Dis. 2008;26(3):272–276. [DOI] [PubMed] [Google Scholar]
- 13.Riddell DR, Graham A, Owen JS. Apolipoprotein E inhibits platelet aggregation through the L-arginine:nitric oxide pathway. Implications for vascular disease. J. Biol. Chem. 1997;272(1):89–95. [DOI] [PubMed] [Google Scholar]
- 14.van der Stoep M, Korporaal SJA, Van Eck M. High-density lipoprotein as a modulator of platelet and coagulation responses. Cardiovasc. Res. 2014;103(3):362–371. [DOI] [PubMed] [Google Scholar]
- 15.van Lier M, Verhoef S, Cauwenberghs S, et al. Role of membrane cholesterol in platelet calcium signalling in response to VWF and collagen under stasis and flow. Thromb. Haemost. 2008;99(6):1068–1078. [DOI] [PubMed] [Google Scholar]
- 16.Cattaneo M, Cerletti C, Harrison P, et al. Recommendations for the Standardization of Light Transmission Aggregometry: A Consensus of the Working Party from the Platelet Physiology Subcommittee of SSC/ISTH. J. Thromb. Haemost. 2013; [DOI] [PubMed] [Google Scholar]
- 17.Splansky GL, Corey D, Yang Q, et al. The Third Generation Cohort of the National Heart, Lung, and Blood Institute’s Framingham Heart Study: design, recruitment, and initial examination. Am. J. Epidemiol. 2007;165(11):1328–1335. [DOI] [PubMed] [Google Scholar]
- 18.Miller M, Stone NJ, Ballantyne C, et al. Triglycerides and cardiovascular disease: A scientific statement from the American Heart Association. Circulation. 2011;123(20):2292–2333. [DOI] [PubMed] [Google Scholar]
- 19.Chan MV, Leadbeater PD, Watson SP, et al. Not all light transmission aggregation assays are created equal: qualitative differences between light transmission and 96-well plate aggregometry. Platelets. 2018;29(7):686–689. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Grech J, Chan MV, Ochin C, et al. Serotonin-Affecting Antidepressant Use in Relation to Platelet Reactivity. Clin. Pharmacol. Ther. 2022;111(4):909–918. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Thalén S, Forsling I, Eintrei J, et al. Pneumatic tube transport affects platelet function measured by multiplate electrode aggregometry. Thromb. Res. 2013;132(1):77–80. [DOI] [PubMed] [Google Scholar]
- 22.Lorenzen H, Frøstrup A-B, Larsen AS, et al. Pneumatic tube transport of blood samples affects global hemostasis and platelet function assays. Int. J. Lab. Hematol. 2021;43(5):1207–1215. [DOI] [PubMed] [Google Scholar]
- 23.Monaco G, Chen H, Poidinger M, et al. flowAI: automatic and interactive anomaly discerning tools for flow cytometry data. Bioinformatics. 2016;32(16):2473–2480. [DOI] [PubMed] [Google Scholar]
- 24.Arima Y, Kaikita K, Ishii M, et al. Assessment of platelet-derived thrombogenicity with the total thrombus-formation analysis system in coronary artery disease patients receiving antiplatelet therapy. J. Thromb. Haemost. 2016;14(4):850–859. [DOI] [PubMed] [Google Scholar]
- 25.R Core Team. A Language and Environment for Statistical Computing. 2021; [Google Scholar]
- 26.Aviram M, Furman B, Brook JG. Chylomicrons from patients with type V hyperlipoproteinemia inhibit platelet function. Atherosclerosis. 1985;56(2):157–167. [DOI] [PubMed] [Google Scholar]
- 27.Naqvi TZ, Shah PK, Ivey PA, et al. Evidence that high-density lipoprotein cholesterol is an independent predictor of acute platelet-dependent thrombus formation. Am. J. Cardiol. 1999;84(9):1011–1017. [DOI] [PubMed] [Google Scholar]
- 28.Osmancik PP, Bednar F, Móciková H. Glycemia, triglycerides and disease severity are best associated with higher platelet activity in patients with stable coronary artery disease. J. Thromb. Thrombolysis. 2007;24(2):105–107. [DOI] [PubMed] [Google Scholar]
- 29.Park B-J, Shim J-Y, Lee H-R, et al. The relationship of platelet count, mean platelet volume with metabolic syndrome according to the criteria of the American Association of Clinical Endocrinologists: a focus on gender differences. Platelets. 2012;23(1):45–50. [DOI] [PubMed] [Google Scholar]
- 30.Michelson AD, Barnard MR, Hechtmant HB, et al. In vivo tracking of platelets : Circulating degranulated platelets rapidly lose surface P-selectin but continue to circulate and function. 1996;93(October):11877–11882. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Kameda S, Sakata T, Kokubo Y, et al. Association of platelet aggregation with lipid levels in the Japanese population: The suita study. J. Atheroscler. Thromb. 2011;18(7):560–567. [DOI] [PubMed] [Google Scholar]
- 32.Ingerman-Wojenski C, Smith JB, Silver MJ. Evaluation of electrical aggregometry: comparison with optical aggregometry, secretion of ATP, and accumulation of radiolabeled platelets. J. Lab. Clin. Med. 1983;101(1):44–52. [PubMed] [Google Scholar]
- 33.Maayani S, Tagliente TM, Schwarz T, et al. The balance of concurrent aggregation and deaggregation processes in platelets is linked to differential occupancy of ADP receptor subtypes. Platelets [electronic article]. 2001;12(2):83–93. ( 10.1080/09537100020032846) [DOI] [PubMed] [Google Scholar]
- 34.Wadowski PP, Eichelberger B, Kopp CW, et al. Disaggregation Following Agonist-Induced Platelet Activation in Patients on Dual Antiplatelet Therapy. J. Cardiovasc. Transl. Res. 2017;10(4):359–367. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Faraday N, Yanek LR, Vaidya D, et al. Leukocyte count is associated with increased platelet reactivity and diminished response to aspirin in healthy individuals with a family history of coronary artery disease. Thromb. Res. 2009;124(3):311–317. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Gofman JW, Jones HB, Lindgren FT, et al. Blood lipids and human atherosclerosis. Circulation. 1950;2(2):161–178. [DOI] [PubMed] [Google Scholar]
- 37.Dyszkiewicz-Korpanty AM, Frenkel EP, Sarode R. Approach to the assessment of platelet function: comparison between optical-based platelet-rich plasma and impedance-based whole blood platelet aggregation methods. Clin. Appl. Thromb. Off. J. Int. Acad. Clin. Appl. Thromb. 2005;11(1):25–35. [DOI] [PubMed] [Google Scholar]
- 38.Aberra T, Peterson ED, Pagidipati NJ, et al. The association between triglycerides and incident cardiovascular disease : What is ‘“ optimal ”‘? J. Clin. Lipidol. [electronic article]. 2020;14(4):438–447.e3. ( 10.1016/j.jacl.2020.04.009) [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Sumner AE, Cowie CC. Ethnic differences in the ability of triglyceride levels to identify insulin resistance. 2008;196:696–703. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.