Skip main navigation

Inactive Matrix Gla-Protein Is Associated With Arterial Stiffness in an Adult Population–Based Study

Originally published 2015;66:85–92


Increased pulse wave velocity (PWV) is a marker of aortic stiffness and an independent predictor of mortality. Matrix Gla-protein (MGP) is a vascular calcification inhibitor that needs vitamin K to be activated. Inactive MGP, known as desphospho-uncarboxylated MGP (dp-ucMGP), can be measured in plasma and has been associated with various cardiovascular markers, cardiovascular outcomes, and mortality. In this study, we hypothesized that high levels of dp-ucMGP are associated with increased PWV. We recruited participants via a multicenter family-based cross-sectional study in Switzerland. Dp-ucMGP was quantified in plasma by sandwich ELISA. Aortic PWV was determined by applanation tonometry using carotid and femoral pulse waveforms. Multiple regression analysis was performed to estimate associations between PWV and dp-ucMGP adjusting for age, renal function, and other cardiovascular risk factors. We included 1001 participants in our analyses (475 men and 526 women). Mean values were 7.87±2.10 m/s for PWV and 0.43±0.20 nmol/L for dp-ucMGP. PWV was positively associated with dp-ucMGP both before and after adjustment for sex, age, body mass index, height, systolic and diastolic blood pressure (BP), heart rate, renal function, low- and high-density lipoprotein, glucose, smoking status, diabetes mellitus, BP and cholesterol lowering drugs, and history of cardiovascular disease (P≤0.01). In conclusion, high levels of dp-ucMGP are independently and positively associated with arterial stiffness after adjustment for common cardiovascular risk factors, renal function, and age. Experimental studies are needed to determine whether vitamin K supplementation slows arterial stiffening by increasing MGP carboxylation.


Increased pulse wave velocity (PWV) is the gold standard in assessing stiffening of the aorta.1 As shown by 3 population studies, all including over thousand participants, PWV refines risk stratification above and beyond traditional cardiovascular risk factors. In multivariable-adjusted analyses, per one standard deviation (SD) increment in PWV, cardiovascular mortality increased by 20% and 60% in a Danish and Taiwanese population samples, respectively.2,3 The Framingham heart study found that one standard deviation increment of PWV was associated with a 48% increase in cardiovascular disease risk.4

Matrix Gla-protein (MGP) is acting as a strong inhibitor of soft tissue calcification.5,6 MGP knockout mice develop massive vascular calcification in their first weeks of life and die within 2 months of vessels rupture.7 MGP is expressed primarily by chondrocytes and vascular smooth muscle cells.8,9 To acquire its full calcification inhibitory activity, MGP needs to undergo 2 post-translational modifications: glutamate carboxylation and serine phosphorylation. Both modifications are not exerted completely, so theoretically, 4 different MGP conformations can be found: unmodified and inactive as desphospho-uncarboxylated MGP (dp-ucMGP), only phosphorylated, only carboxylated, and finally fully modified and active as phosphorylated and carboxylated MGP.10,11 Carboxylation of MGP is a vitamin K–dependent process: vitamin K hydroquinone, a vitamin K metabolite is an active cofactor of the carboxylation of glutamate residues into gammacarboxyglutamate.12 Active MGP adsorbs to extracellular hydroxyapatite, thus inhibiting crystal growth.13 It also inhibits vascular smooth muscle cell apoptosis by binding to bone morphogenetic protein 2, a proapoptotic agonist.14 Two clinical studies investigated the link between dp-ucMGP and CT-detected calcifications. High dp-ucMGP levels were associated with below-knee arterial calcification score among type 2 diabetic patients.15 A trend in an association between higher circulating dp-ucMGP levels and high coronary artery calcification was observed in postmenopausal healthy women (P=0.06).16

The objective of this study was to evaluate whether high levels of dp-ucMGP are associated with increased vascular stiffness as measured by aortic PWV in a random adult population sample recruited within the framework of the Swiss Kidney Project on Genes in Hypertension (SKIPOGH).


SKIPOGH is a family-based cross-sectional study exploring the role of genes and kidney hemodynamics in BP regulation and kidney function in the general adult population.17

Study Visits and Definitions

Participants were seen in the morning after an overnight fast. BP was measured 5 times in clinic after a resting period of at least 10 minutes in the sitting position with an appropriately sized arm cuff and a nonmercury manual auscultatory sphygmomanometer (A&D UM-101; A&D Company Ltd, Toshima Ku, Tokyo, Japan),18 according to the guidelines of the European Society of Hypertension. BP and heart rate were the mean of 4 measurements before the recording of PWV. Diabetes mellitus was a self-reported diagnosis or being on treatment with antidiabetic drugs. Cardiovascular disease was considered positive if participants reported a history of stroke or heart disease.

Pulse Wave Velocity

The arterial waveform was assessed during an 8 second period in the supine position after 15 minutes of rest at the carotid and femoral arteries by applanation tonometry, using a high-fidelity SPC-301 micromanometer (Millar Instruments, Inc., Houston), interfaced with a laptop computer running the SphygmoCor software, version 8.0 or 8.2 (AtCor Medical Pty. Ltd., West Ryde, Australia).

In each center, a single board certified nephrologist (B. Ponte, M. Pruijm, D. Ackermann) obtained the arterial waveforms. Using a tape, they measured the distance from the suprasternal notch to the carotid sampling site (distance A) and the distance from the suprasternal notch to the femoral sampling site (distance B). Pulse wave travel distance was distance B minus distance A. Pulse transit time was derived from the average of 10 consecutive beats. Carotid–femoral PWV is the ratio of travel distance in meters to transit time in seconds.

We assessed intraobserver and interobserver reproducibility in 40 and 15 participants, respectively, according to Bland and Altman’s method.19 The intraobserver and interobserver biases averaged 0.27 m/s (95% confidence interval, 0.32–0.85) and 0.43 m/s (95% confidence interval, 0.05–0.82), respectively (Figure S1 in the online-only Data Supplement).

Laboratory Procedures

Venous blood samples were drawn after an overnight fast. Kidney function, serum cholesterol, and glucose were analyzed by standard clinical laboratory methods. Dp-ucMGP was assessed in EDTA plasma using the inaKtif MGP iSYS kit (IDS, Boldon, UK), which is a precommercial dual-antibody test based on the sandwich ELISA developed by VitaK, Maastricht University, The Netherlands. Intra-assay variation coefficients are 3.1% for the upper limit of the normal range and 5.4% for the lower limit of the normal range. Inter-assay variation coefficients are 6.9% for the upper limit of the normal range and 13.6% for the lower limit of the normal range (unpublished data). Chronic Kidney Disease Epidemiology Collaboration formula was used to estimate the glomerular filtration rate using serum creatinine.20

Study Participants

From December 2009 to March 2013, 1128 adult participants were recruited in 2 regions (Berne and Geneva) and one city (Lausanne) of Switzerland. A random sample of the inhabitants was drawn using different strategies. Inclusion criteria were (1) having a minimum age of 18 years; (2) being of European ancestry; (3) having at least 1 and ideally 3 first-degree family members willing to participate; and (4) willingness to provide written informed consent. Pregnant or breastfeeding women were not included. The participation rate was 25.6%. The SKIPOGH study was approved by the Institutional Ethical Review Boards of each of the participating university hospitals. Twelve participants had no or insufficient plasma, so 1116 participants in the SKIPOGH study had dp-ucMGP measurements. We excluded 3 participants taking vitamin K antagonists. Ninety-four participants had no PWV measurements for various reasons (obesity, arrhythmias, technical problems, no show). Eighteen participants were removed because of extreme outliers (Table S1) defined as observations lying >3 interquartile ranges away from percentile 25 and 75, leaving 1001 (90%) participants for further analysis. The number of participants at each step is shown in Figure 1.

Figure 1.

Figure 1. Flowchart of Participants. dp-ucMGP indicates desphospho-uncarboxylated matrix gla-protein; PWV, pulse wave velocity; and SKIPOGH, Swiss Kidney Project on Genes in Hypertension.


Statistical analyses were conducted using STATA® version 13 (StataCorp, College Station, TX). Continuous and categorical variables were expressed as mean±SD and as number of participants (%), respectively. Statistical significance was a 2-sided P value of <0.05. Analysis of variance was used to compare variables between tertiles of dp-ucMGP. Medians of dp-ucMGP and PWV were used to separate participants into 4 groups. χ2 test of independence was applied to look for an association between the 2 variables, and the odds ratio was calculated. We conducted univariate analyses to look for associations between PWV and the following variables of interest: dp-ucMGP, sex, age, body mass index, height, systolic and diastolic pressure, heart rate, renal function (estimated glomerular filtration rate), low- and high-density lipoprotein, smoking status, diabetes mellitus, BP and cholesterol lowering drugs, and history of cardiovascular disease. The technique of multiple imputation was used to estimate missing values for the above-mentioned independent variables, except for dp-ucMGP. We applied multivariable linear mixed regression to determine the independent association of PWV with each variable, with family clusters modeled as a random effect. All models were also adjusted for centers as a fixed effect. Using a backward elimination procedure, removing one by one the variable with the highest nonsignificant P value, only significant variables (P>0.05) were kept in the final model. For continuous variables, results of all linear analyses are presented as β coefficients per standard deviation and associated P values. For continuous variables, β coefficients represent the increase in PWV per one standard deviation increment of the corresponding independent variable. For categorical variables, β coefficients represent the difference in PWV between the 2 categories.


Mean values were 7.87±2.10 m/s for PWV and 0.43±0.20 nmol/L for dp-ucMGP. 526 (53%) participants were women. Age averaged (±SD) 46.5±17.2 years, body mass index 24.7±4.2 kg/m2, height 1.71±0.09 m, BP 117.0±16.4 mm Hg systolic and 75.3±9.6 mm Hg diastolic, and heart rate 66.4±10.5 beats per minute. Renal function (estimated glomerular filtration rate) averaged 97.0±17.4 mL/min per 1.73 m2, low- and high-density lipoprotein 3.12±0.93 and 1.52±0.42 mmol/L, respectively. Glucose mean was 5.15±0.68 mmol/L. Among all participants, 247 (25%) were smoker, 24 (2.4%) had diabetes mellitus, 146 (14.6%) and 37 (3.7%) were on BP or cholesterol lowering drugs, respectively. Ninety-nine (9.9%) participants reported a history of cardiovascular disease. Population characteristics are described in Table 1 according to tertiles of dp-ucMGP and in Table S2 according to sex. Across increasing categories of dp-ucMGP, more participants had diabetes, were on BP lowering drugs, or had a history of cardiovascular disease, whereas fewer reported smoking. Furthermore, age, body mass index, BP, low density lipoprotein and glucose increased with higher category of dp-ucMGP (all P<0.01). Renal function and high-density lipoprotein decreased with higher category of dp-ucMGP (all P<0.01).

Table 1. Baseline Characteristics of Participants by Tertiles of the dp-ucMGP Distribution

CharacteristicLowMediumHighP ValueTotal
dp-ucMGP limits, nmol/L<0.320.32–0.50>0.50
N with characteristics, %
 Women187 (53.9)176 (51.3)163 (52.4)0.79526 (52.5)
 Current smoker112 (32.3)84 (24.5)51 (16.4)<0.01247 (24.7)
 Diabetes mellitus6 (1.7)6 (1.8)14 (4.5)0.0426 (2.6)
 BP lowering drugs23 (6.6)46 (13.4)77 (24.7)<0.01146 (14.6)
 Cholesterol lowering drugs8 (2.3)12 (3.5)17 (5.4)0.0937 (3.7)
 CV disease20 (5.8)33 (9.6)46 (14.7)<0.0199 (9.9)
Mean of characteristic (±SD)
 PWV, m/s6.99±1.397.75±2.008.89±2.36<0.017.87±2.10
 Age, y38.7±14.945.9±16.755.0±16.1<0.0146.5±17.2
 Body mass index, kg/m222.9±3.224.5±4.126.7±4.4<0.0124.7±4.2
 Height, m1.71±0.091.72±0.091.69±±0.09
 Systolic BP, mm Hg111±13117±16124±18<0.01117±16
 Diastolic BP, mm Hg73±976±1078±9.0<0.0175±10
 Heart rate, bpm66.5±11.165.8±10.167.0±10.10.6266.4±10.5
 eGFR, mL/min per 1.73 m2105±1498±1788±18<0.0197±17
 LDL, mmol/L2.90±0.893.12±0.923.33±0.93<0.013.12±0.93
 HDL, mmol/L1.60±0.441.48±0.411.49±0.41<0.011.52±0.42
 Glucose, mmol/L4.97±0.595.10±0.575.38±0.80<0.015.15±0.68
dp-ucMGP, nmol/L0.23±0.070.41±0.050.66±0.15<0.010.43±0.20

P-values are for linear trend across tertiles of the dp-ucMGP distribution. Diabetes mellitus was a self-reported diagnosis or being on treatment with antidiabetic drugs. CV disease was considered positive if participants reported a history of stroke or heart disease. BP indicates blood pressure; CV, cardiovascular; Dp-ucMGP, desphospho-uncarboxylated matrix gla-protein; eGFR, estimated glomerular filtration rate using the Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) formula; HDL, high density lipoprotein; LDL, low density lipoprotein; and PWV, pulse wave velocity.

In univariate analyses, PWV was significantly associated with all independent variables except heart rate (Table 3). PWV was positively associated with dp-ucMGP (Figure 2); a one SD increase in dp-ucMGP (0.20 nmol/L) was associated with a 0.9 m/s increase in PWV. PWV was positively associated with age, body mass index, low-density lipoprotein, glucose, systolic and diastolic BPs (all P<0.01). It was negatively associated with height (P<0.01), renal function (P<0.01), and high-density lipoprotein (P=0.03). Participants on BP and cholesterol lowering drugs and with diabetes mellitus or cardiovascular disease had a higher PWV (all P<0.01). Current smokers and women had a lower PWV. Distribution of participants according to dp-ucMGP and PWV medians is shown in Table 2. The odd ratio for a PWV >7.4 (m/s) was 3.25 (95% CI, 2.51; 4.21) for higher compared with lower dp-ucMGP levels (≤0.408 nmol/L).

Table 2. Distribution of Participants According to dp-ucMGP and PWV Medians

VariablesDp-ucMGP, nmol/LTotal
Low (≤0.408)High (>0.408)
PWV, m/s
 Low (≤7.4)324179503
 High (>7.4)178320498

Odd ratio 3.25 (95% CI, 2.51; 4.21), Pearson’s chi-squared =82.28, P-value <0.001. CI indicates confidence interval; Dp-ucMGP, desphospho-uncarboxylated matrix gla-protein; and PWV, pulse wave velocity.

Figure 2.

Figure 2. Scatterplot of pulse wave velocity (PWV; m/s) and desphospho-uncarboxylated matrix gla-protein (dp-ucMGP; nmol/L).

In multivariate analysis (Table 3), we identified the following independent variables as significant positive determinants of PWV: dp-ucMGP, age, body mass index, systolic pressure, heart rate, glucose, diabetes mellitus, and history of cardiovascular disease. Diastolic pressure was a significant negative determinant of PWV. Heart rate was forced into the model for clinical coherence, despite a nonsignificant P value in the univariate analysis.21 During the backward elimination procedure, independent variables were removed in the following order: HDL, LDL, current smoker, estimated glomerular filtration rate, cholesterol lowering drugs, sex, BP lowering drugs, and finally height (Table S3). Heart rate remained significant in each step of the procedure. In the fully adjusted model, a one SD increase in dp-ucMGP (0.20 nmol/L) was associated with a 0.2 m/s increase in PWV. Beta coefficients are displayed in Table 3 for univariate models, the full model, and the model with only significant independent variables and heart rate. Analysis with pulse pressure and mean pressure (Table S4), with centers as a random effect (Table S5), with extreme outliers (Table S6), with an interaction between age and dp-ucMGP (Table S7) and sensitivity analysis (Table S8 and S9), are presented in the online-only Data Supplement. These additional analyses led to similar results and did not alter our conclusions.

Table 3. Factors Associated With PWV in Univariate and Multivariate Linear Mixed Regression Models (N=1001)

Independent VariablesSDUnivariate ModelsFull Model*Model With Significant Independent Variables
βP ValueβP ValueβP Value95% Confidence Intervals
Dp-ucMGP, nmol/L0.200.901<0.0010.186<0.0010.198<0.001§0.1110.277
Age, y17.21.531<0.0010.977<0.0010.979<0.0010.8811.135
Body mass index, kg/m24.220.810<0.0010.109<0.0330.1040.0290.0110.224
Height, m0.09−0.278<0.0010.1100.066
Systolic pressure, mm Hg16.41.361<0.0010.743<0.0010.746<0.0010.6200.873
Diastolic pressure, mm Hg9.60.854<0.001−0.1630.009−0.1480.013−0.265−0.031
Heart rate, bpm10.50.1050.1100.1210.0050.1050.0110.0240.186
eGFR, mL/min per 1.73 m217.4-1.166<0.001−0.0330.589
LDL, mmol/L0.930.492<0.001−0.0080.866
HDL, mmol/L0.42−0.1490.027−0.0060.895
Glucose, mmol/L0.680.964<0.0010.1620.0020.1650.0010.0650.256
Sex (women)−0.2820.0330.1050.408
Current smoker (yes)−0.683<0.001−0.0180.850
Diabetes mellitus (yes)3.207<0.0010.8020.0040.7920.0040.2591.325
BP lowering drugs (yes)2.463<0.0010.1880.145
Cholesterol lowering drugs (yes)1.861<0.001−0.1650.445
CV disease (yes)1.757<0.0010.3010.0320.3410.0140.0700.612

All models are adjusted for the center where the PWV was recorded, as a fixed effect. Diabetes mellitus was a self-reported diagnosis or being on treatment with antidiabetic drugs. CV disease was considered positive if participants reported a history of stroke or heart disease. BP indicates blood pressure; CV, cardiovascular; dp-ucMGP, desphospho-uncarboxylated matrix gla-protein; eGFR, estimated glomerular filtration rate using the Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) formula; HDL, high density lipoprotein; LDL, low density lipoprotein; and PWV, pulse wave velocity.

*Heart rate was not significant in univariate analysis but forced in the full model for clinical coherence.

Model reached at the end of the backward procedure, all steps are presented in Table S3.

β, beta coefficient; for continuous variables it represents the increase in PWV per an increase of one standard deviation (SD) of the corresponding independent variable and for categorical variables, the difference of PWV between the 2 categories.

§P value <0.00001.


In this multicentric population-based study, PWV was positively associated with dp-ucMGP in 1001 adults of European descent, independently of other cardiovascular risk factors.

To our knowledge, there is no published study that has found a significant association between arterial stiffness and any MGP species in the general adult population. Some studies in selected groups of patients used total MGP or uncarboxylated MGP (ucMGP), without taking phosphorylation status into account, and found no significant association with arterial stiffness. In a study conducted in 144 postmenopausal women, PWV was not associated with total MGP, but was associated with osteoprotegerin, another potential vascular calcification inhibitor.22 In 81 renal transplant recipients, PWV was again not associated with total MGP.23 In 120 dialysis patients, PWV was not associated this time with ucMGP.24 The authors hypothesized a possible sequestration of ucMGP near the calcified areas of arteries because ucMGP levels were also lower in dialysis patients than in healthy control. In 61 children on dialysis, PWV was not associated with ucMGP levels, yet levels were also lower in children on dialysis than in healthy controls.25 As the molecular weight of MGP is 14 kDa, those lower levels cannot be the consequence of the dialysis itself.26 Dp-ucMGP has a low affinity for calcium in the absence of phosphoserine and gammacarboxyglutamate residues13 and is probably prone to be set free in the circulation. Moreover, it is elevated in patients at risk of cardiovascular calcification because of end-stage renal disease.27 Finding a significant association between PWV and dp-ucMGP is a major step forward after all those unsuccessful attempts. It is also a further argument to state that dp-ucMGP probably best reflects MGP bioactivity rather than other MGP species.

Whether the observed association reflects a causal effect of MGP on arterial stiffness, or whether MGP is merely a biomarker of vascular vitamin K status, is unclear. We hypothesize that MGP lies on the causal pathway between arterial stiffness and vitamin K status. Vitamin K influences MGP carboxylation. Two randomized double-blind, placebo-controlled intervention trials bring forward some arguments in favor of a causal role of vitamin K status on arterial stiffness. In the first one, after 3 years of vitamin K1 supplementation, elastic properties of the common carotid artery remained similar in the supplementation group, whereas they worsened in the placebo group.28 In the vitamin K1 group, the percentage change from baseline for the distensibility coefficient and the compliance coefficient were −0.8% and +2.7%, respectively. In the placebo group, they were −9.6% and −5.9%, respectively. The differences between the 2 groups were significant (P<0.05). The daily supplementation dose of vitamin K1 used (1 mg) was around 10× the recommended dose. The second trial used 180 μg of vitamin K2 also for 3 years. PWV and dp-ucMGP were measured, among other determinants of vessel wall properties. The results showed a trend in a reduction of PWV (P=0.09) and a significant reduction of dp-ucMGP in the vitamin K2 group.29 Pearson correlation between baseline PWV and dp-ucMGP was significant at 0.16 (P<0.05), and no multivariable analysis was performed. Participants in both studies were healthy postmenopausal women: 181 in the first trial and 244 in the second one. In a case–control study with a mean follow up of 1.2 years, 18 hemodialysis patients on warfarin had a greater increase in PWV than controls not on warfarin.30 Controls were age-matched (±2 years) and also on hemodialysis. All participants were free of previous cardiovascular disease. Adjusted PWV increased by 0.86 m/s in control and by 2.24 m/s in warfarin group; the difference was significant (P=0.013), despite already high mean PWV at baseline in both groups (12.8 m/s). Finally, a recent Mendelian randomization study found that higher dp-ucMGP predicts noncancer mortality.31 The Mendelian design strongly supports a causal relationship and also favors the hypothesis that dp-ucMGP is not solely a biomarker of vascular vitamin K status, although both hypotheses could be true.

In healthy arteries, most of the MGP occurs in the carboxylated (active) form, and it is found in the tunica media along the elastin fibers.32 A potential mechanism by which MGP could influence arterial stiffness is via decreased vascular calcifications, considering the known relation between arterial stiffness and the degree of vascular calcification. In rats, deposition of calcium on the medial elastic fibers is responsible for aortic stiffness and an increase in PWV.33 In the aortic media, calcifications are close to, and stiffen, elastic fibers.34 Although calcium can be deposited into either the medial or intimal layers of the vasculature, experts hold the view that medial calcifications, rather than intimal calcifications, influence arterial stiffness.35 Medial calcification is particularly common in diabetes mellitus and advanced renal disease,36 and uncarboxylated (inactive) MGP is found near calcifications.32 Circulating levels of dp-ucMGP do not affect vascular calcifications because dp-ucMGP is biologically inactive and not able to bind calcification and bone morphogenetic protein 2.37,38 However, low levels, as seen with vitamin K supplementation, reflect strong MGP carboxylation and so maximum inhibition of calcification in the arterial wall.27 Two ongoing randomized trials currently explore whether vitamin K supplementation decreases arterial and valvular calcifications in hemodialysis patients.39 The results of cardiovascular prevention trials in the hemodialysis population are often controversial, such as with respect to lipid lowering approaches.40 Irrespective of the outcomes of the ongoing studies in hemodialysis patients, our findings remain of great importance in a general population level. Given the absence of known toxicity of vitamin K supplementation, our data might lead to a large scale intervention in a lower risk population. PWV is an easy, noninvasive, vascular assessment in any future study on the benefits of a primary and earlier supplementation. Furthermore, dp-ucMGP levels enable to monitor compliance on supplementation, adapt vitamin K doses to nutritional intake of the individual, select the best vitamin K type (K1, K2), and choose the right participants. Selecting participants with high dp-ucMGP levels is expected to maximize any benefit and statistical power.

In vitro studies and animal models convincingly demonstrated that MGP functions as a vascular calcification inhibitor.7,10 Humans with absent or nonfunctional MGP develop Keutel syndrome, an autosomal recessive disorder characterized by abnormal cartilage calcification, peripheral pulmonary stenosis, and midfacial hypoplasia.41 Autopsy of one young suspected Keutel patient, not confirmed by genetic analysis, has shown extensive arterial calcification.42 By contrast, after a 26-year follow-up, there was no evidence of vascular calcification on chest and abdominal CT in 4 genetically confirmed women with Keutel syndrome.43 Similarly, cranial CT scan and multislice CT scanning of the coronary arteries revealed no calcification in a 21-year-old men with genetically confirmed Keutel syndrome.44 Unfortunately, their PWVs were not reported. One hypothesis that could be in line with our results and the absence of calcifications in tomography in Keutel patients is that a high level of inactive MGP leads to an increase in the number, or in the size, of calcifications, but sufficiently scattered in the vessels not to be detectable by tomography. Because to be taken into account in radiological calcification scores, a calcific lesion has to be ≥1 mm2 and with a tomographic density of at least 130 Hounsfield units.45,46

The limitations of the present study include the observational nature of the association, which limits causal inference. We also lack direct assessments of vascular calcification. In our current study, dp-ucMGP was measured at a single time point. However, previous studies showed that dp-ucMGP levels remain unchanged over a 12-week interval.47 High dp-ucMGP levels could result from an increased synthesis rather than insufficient carboxylation because of lack of vitamin K. Yet, the fact that vitamin K (MK-4) supplementation significantly and massively decreases dp-ucMGP allows us to conclude that it would represent only a small fraction of circulating dp-ucMGP.27 The participation rate was low mainly because the goal was to enroll at least 3 first-degree family members, yet age and sex distributions are similar to that of the adult Swiss population (Table S10). Although we cannot exclude some level of residual confounding, it is unlikely that unmeasured confounders would eliminate or reverse the observed PWV–dp-ucMGP association. Finally, PWV examinations were performed by different operators in different centers. However, linear models were adjusted for the examination centers. Moreover, intra- and interobserver variability showed an excellent reproducibility of the estimates of PWV.

Our results shed new light on the role of MGP in arterial stiffness in humans. Because vitamin K therapy lowers dp-ucMGP levels,27 these results are a further incentive to conduct additional trials to study the impact of vitamin K supplementation on arterial calcifications and stiffness, in particular in patients at high cardiovascular risk.


High levels of dp-ucMGP are associated with increased arterial stiffness, measured by PWV, independently of common cardiovascular risk factors in the general adult population. Considering that dp-ucMGP is a marker of vascular vitamin K status, our results are in line with experimental studies in humans showing that vitamin K supplementation stabilizes vascular arterial properties.28 Randomized clinical trials are needed to further explore whether vitamin K supplementation reduces cardiovascular risk.


We are extremely grateful to the SKIPOGH (Swiss Kidney Project on Genes in Hypertension) study participants. We thank the study nurses involved in the study and recruitment: Marie-Odile Levy, Guler Gök-Sogüt, Ulla Spüchbach, and Dominique Siminski. We thank Sandrine Estoppey for her help in the logistic and database management.


*These authors contributed equally to this work.

This article was sent to Marc L. De Buyzere, Guest Editor, for review by expert referees, editorial decision, and final disposition.

The online-only Data Supplement is available with this article at

Correspondence to Murielle Bochud, Division des Maladies Chroniques, Institut Universitaire de Médecine Sociale et Préventive, Route de la Corniche 10, CH-1010 Lausanne. E-mail


  • 1. O’Rourke MF, Staessen JA, Vlachopoulos C, Duprez D, Plante GE.Clinical applications of arterial stiffness; definitions and reference values.Am J Hypertens. 2002; 15:426–444.CrossrefMedlineGoogle Scholar
  • 2. Willum-Hansen T, Staessen JA, Torp-Pedersen C, Rasmussen S, Thijs L, Ibsen H, Jeppesen J.Prognostic value of aortic pulse wave velocity as index of arterial stiffness in the general population.Circulation. 2006; 113:664–670. doi: 10.1161/CIRCULATIONAHA.105.579342.LinkGoogle Scholar
  • 3. Wang KL, Cheng HM, Sung SH, Chuang SY, Li CH, Spurgeon HA, Ting CT, Najjar SS, Lakatta EG, Yin FC, Chou P, Chen CH.Wave reflection and arterial stiffness in the prediction of 15-year all-cause and cardiovascular mortalities: a community-based study.Hypertension. 2010; 55:799–805. doi: 10.1161/HYPERTENSIONAHA.109.139964.LinkGoogle Scholar
  • 4. Mitchell GF, Hwang SJ, Vasan RS, Larson MG, Pencina MJ, Hamburg NM, Vita JA, Levy D, Benjamin EJ.Arterial stiffness and cardiovascular events: the Framingham Heart Study.Circulation. 2010; 121:505–511. doi: 10.1161/CIRCULATIONAHA.109.886655.LinkGoogle Scholar
  • 5. Shanahan CM, Proudfoot D, Farzaneh-Far A, Weissberg PL.The role of Gla proteins in vascular calcification.Crit Rev Eukaryot Gene Expr. 1998; 8:357–375.CrossrefMedlineGoogle Scholar
  • 6. Price PA, Urist MR, Otawara Y.Matrix Gla protein, a new gamma-carboxyglutamic acid-containing protein which is associated with the organic matrix of bone.Biochem Biophys Res Commun. 1983; 117:765–771.CrossrefMedlineGoogle Scholar
  • 7. Luo G, Ducy P, McKee MD, Pinero GJ, Loyer E, Behringer RR, Karsenty G.Spontaneous calcification of arteries and cartilage in mice lacking matrix GLA protein.Nature. 1997; 386:78–81. doi: 10.1038/386078a0.CrossrefMedlineGoogle Scholar
  • 8. Shanahan CM, Cary NR, Metcalfe JC, Weissberg PL.High expression of genes for calcification-regulating proteins in human atherosclerotic plaques.J Clin Invest. 1994; 93:2393–2402. doi: 10.1172/JCI117246.CrossrefMedlineGoogle Scholar
  • 9. Shanahan CM, Weissberg PL, Metcalfe JC.Isolation of gene markers of differentiated and proliferating vascular smooth muscle cells.Circ Res. 1993; 73:193–204.LinkGoogle Scholar
  • 10. Price PA, Faus SA, Williamson MK.Warfarin causes rapid calcification of the elastic lamellae in rat arteries and heart valves.Arterioscler Thromb Vasc Biol. 1998; 18:1400–1407.LinkGoogle Scholar
  • 11. Price PA, Rice JS, Williamson MK.Conserved phosphorylation of serines in the Ser-X-Glu/Ser(P) sequences of the vitamin K-dependent matrix Gla protein from shark, lamb, rat, cow, and human.Protein Sci. 1994; 3:822–830. doi: 10.1002/pro.5560030511.CrossrefMedlineGoogle Scholar
  • 12. Wallin R, Cain D, Sane DC.Matrix Gla protein synthesis and gamma-carboxylation in the aortic vessel wall and proliferating vascular smooth muscle cells–a cell system which resembles the system in bone cells.Thromb Haemost. 1999; 82:1764–1767.CrossrefMedlineGoogle Scholar
  • 13. Goiko M, Dierolf J, Gleberzon JS, Liao Y, Grohe B, Goldberg HA, de Bruyn JR, Hunter GK.Peptides of Matrix Gla protein inhibit nucleation and growth of hydroxyapatite and calcium oxalate monohydrate crystals.PLoS One. 2013; 8:e80344. doi: 10.1371/journal.pone.0080344.CrossrefMedlineGoogle Scholar
  • 14. Boström K.Insights into the mechanism of vascular calcification.Am J Cardiol. 2001; 88(2A):20E–22E.CrossrefMedlineGoogle Scholar
  • 15. Liabeuf S, Bourron O, Olivier B, Vemeer C, Theuwissen E, Magdeleyns E, Aubert CE, Brazier M, Mentaverri R, Hartemann A, Massy ZA.Vascular calcification in patients with type 2 diabetes: the involvement of matrix Gla protein.Cardiovasc Diabetol. 2014; 13:85. doi: 10.1186/1475-2840-13-85.CrossrefMedlineGoogle Scholar
  • 16. Dalmeijer GW, van der Schouw YT, Vermeer C, Magdeleyns EJ, Schurgers LJ, Beulens JW.Circulating matrix Gla protein is associated with coronary artery calcification and vitamin K status in healthy women.J Nutr Biochem. 2013; 24:624–628. doi: 10.1016/j.jnutbio.2012.02.012.CrossrefMedlineGoogle Scholar
  • 17. Pruijm M, Ponte B, Ackermann D, Vuistiner P, Paccaud F, Guessous I, Ehret G, Eisenberger U, Mohaupt M, Burnier M, Martin PY, Bochud M.Heritability, determinants and reference values of renal length: a family-based population study.Eur Radiol. 2013; 23:2899–2905. doi: 10.1007/s00330-013-2900-4.CrossrefMedlineGoogle Scholar
  • 18. Pruijm MT, Wuerzner G, Glatz N, Alwan H, Ponte B, Ackermann D, Burnier M, Bochud M.A new technique for simultaneous validation of two manual nonmercury auscultatory sphygmomanometers (A&D UM-101 and Accoson Greenlight 300) based on the International protocol.Blood Press Monit. 2010; 15:322–325. doi: 10.1097/MBP.0b013e32833f56a8.CrossrefMedlineGoogle Scholar
  • 19. Bland JM, Altman DG.Statistical methods for assessing agreement between two methods of clinical measurement.Lancet. 1986; 1:307–310.CrossrefMedlineGoogle Scholar
  • 20. Levey AS, Stevens LA, Schmid CH, Zhang YL, Castro AF, Feldman HI, Kusek JW, Eggers P, Van Lente F, Greene T, Coresh J; CKD-EPI (Chronic Kidney Disease Epidemiology Collaboration). A new equation to estimate glomerular filtration rate.Ann Intern Med. 2009; 150:604–612.CrossrefMedlineGoogle Scholar
  • 21. Liu YP, Richart T, Li Y, Zhan WW, Staessen JA.Is arterial stiffness related to body height?Hypertension. 2010; 55:e24–e25. doi: 10.1161/HYPERTENSIONAHA.110.152553.LinkGoogle Scholar
  • 22. Albu A, Fodor D, Bondor C, Crăciun AM.Bone metabolism regulators and arterial stiffness in postmenopausal women.Maturitas. 2013; 76:146–150. doi: 10.1016/j.maturitas.2013.07.001.CrossrefMedlineGoogle Scholar
  • 23. Gungor O, Kircelli F, Carrero JJ, Hur E, Demirci MS, Asci G, Toz H.The effect of immunosuppressive treatment on arterial stiffness and matrix Gla protein levels in renal transplant recipients.Clin Nephrol. 2011; 75:491–496.CrossrefMedlineGoogle Scholar
  • 24. Hermans MM, Vermeer C, Kooman JP, Brandenburg V, Ketteler M, Gladziwa U, Rensma PL, Leunissen KM, Schurgers LJ.Undercarboxylated matrix GLA protein levels are decreased in dialysis patients and related to parameters of calcium-phosphate metabolism and aortic augmentation index.Blood Purif. 2007; 25:395–401. doi: 10.1159/000108629.CrossrefMedlineGoogle Scholar
  • 25. Shroff RC, Shah V, Hiorns MP, Schoppet M, Hofbauer LC, Hawa G, Schurgers LJ, Singhal A, Merryweather I, Brogan P, Shanahan C, Deanfield J, Rees L.The circulating calcification inhibitors, fetuin-A and osteoprotegerin, but not matrix Gla protein, are associated with vascular stiffness and calcification in children on dialysis.Nephrol Dial Transplant. 2008; 23:3263–3271. doi: 10.1093/ndt/gfn226.CrossrefMedlineGoogle Scholar
  • 26. Schurgers LJ, Uitto J, Reutelingsperger CP.Vitamin K-dependent carboxylation of matrix Gla-protein: a crucial switch to control ectopic mineralization.Trends Mol Med. 2013; 19:217–226. doi: 10.1016/j.molmed.2012.12.008.CrossrefMedlineGoogle Scholar
  • 27. Cranenburg EC, Koos R, Schurgers LJ, Magdeleyns EJ, Schoonbrood TH, Landewé RB, Brandenburg VM, Bekers O, Vermeer C.Characterisation and potential diagnostic value of circulating matrix Gla protein (MGP) species.Thromb Haemost. 2010; 104:811–822. doi: 10.1160/TH09-11-0786.CrossrefMedlineGoogle Scholar
  • 28. Braam LA, Hoeks AP, Brouns F, Hamulyák K, Gerichhausen MJ, Vermeer C.Beneficial effects of vitamins D and K on the elastic properties of the vessel wall in postmenopausal women: a follow-up study.Thromb Haemost. 2004; 91:373–380. doi: 10.1160/TH03-07-0423.CrossrefMedlineGoogle Scholar
  • 29. Knapen MH, Braam LA, Drummen NE, Bekers O, Hoeks AP, Vermeer C.Menaquinone-7 supplementation improves arterial stiffness in healthy postmenopausal women: double-blind randomised clinical trial.Thromb Haemost. 2015; 113:. doi: 10.1160/TH14-08-0675.CrossrefGoogle Scholar
  • 30. Mac-Way F, Poulin A, Utescu MS, De Serres SA, Marquis K, Douville P, Desmeules S, Larivière R, Lebel M, Agharazii M.The impact of warfarin on the rate of progression of aortic stiffness in hemodialysis patients: a longitudinal study.Nephrol Dial Transplant. 2014; 29:2113–2120. doi: 10.1093/ndt/gfu224.CrossrefMedlineGoogle Scholar
  • 31. Liu YP, Gu YM, Thijs L, et al. Inactive matrix Gla protein is causally related to adverse health outcomes: a Mendelian randomization study in a Flemish population.Hypertension. 2015; 65:463–470. doi: 10.1161/HYPERTENSIONAHA.114.04494.LinkGoogle Scholar
  • 32. Schurgers LJ, Teunissen KJ, Knapen MH, Kwaijtaal M, van Diest R, Appels A, Reutelingsperger CP, Cleutjens JP, Vermeer C.Novel conformation-specific antibodies against matrix gamma-carboxyglutamic acid (Gla) protein: undercarboxylated matrix Gla protein as marker for vascular calcification.Arterioscler Thromb Vasc Biol. 2005; 25:1629–1633. doi: 10.1161/01.ATV.0000173313.46222.43.LinkGoogle Scholar
  • 33. Niederhoffer N, Lartaud-Idjouadiene I, Giummelly P, Duvivier C, Peslin R, Atkinson J.Calcification of medial elastic fibers and aortic elasticity.Hypertension. 1997; 29:999–1006.LinkGoogle Scholar
  • 34. Greenwald SE.Ageing of the conduit arteries.J Pathol. 2007; 211:157–172. doi: 10.1002/path.2101.CrossrefMedlineGoogle Scholar
  • 35. Mackey RH, Venkitachalam L, Sutton-Tyrrell K.Calcifications, arterial stiffness and atherosclerosis.Adv Cardiol. 2007; 44:234–244. doi: 10.1159/000096744.CrossrefMedlineGoogle Scholar
  • 36. Proudfoot D, Shanahan CM, Weissberg PL.Vascular calcification: new insights into an old problem.J Pathol. 1998; 185:1–3.CrossrefMedlineGoogle Scholar
  • 37. Murshed M, Schinke T, McKee MD, Karsenty G.Extracellular matrix mineralization is regulated locally; different roles of two gla-containing proteins.J Cell Biol. 2004; 165:625–630. doi: 10.1083/jcb.200402046.CrossrefMedlineGoogle Scholar
  • 38. Malhotra R, Burke MF, Martyn T, et al. Inhibition of bone morphogenetic protein signal transduction prevents the medial vascular calcification associated with matrix Gla protein deficiency.PLoS One. 2015; 10:e0117098. doi: 10.1371/journal.pone.0117098.CrossrefMedlineGoogle Scholar
  • 39. Gallieni M, Fusaro M.Vitamin K and cardiovascular calcification in CKD: is patient supplementation on the horizon?Kidney Int. 2014; 86:232–234. doi: 10.1038/ki.2014.24.CrossrefMedlineGoogle Scholar
  • 40. Palmer SC, Craig JC, Navaneethan SD, Tonelli M, Pellegrini F, Strippoli GF.Benefits and harms of statin therapy for persons with chronic kidney disease: a systematic review and meta-analysis.Ann Intern Med. 2012; 157:263–275. doi: 10.7326/0003-4819-157-4-201208210-00007.CrossrefMedlineGoogle Scholar
  • 41. Munroe PB, Olgunturk RO, Fryns JP, Van Maldergem L, Ziereisen F, Yuksel B, Gardiner RM, Chung E.Mutations in the gene encoding the human matrix Gla protein cause Keutel syndrome.Nat Genet. 1999; 21:142–144. doi: 10.1038/5102.CrossrefMedlineGoogle Scholar
  • 42. Meier M, Weng LP, Alexandrakis E, Rüschoff J, Goeckenjan G.Tracheobronchial stenosis in Keutel syndrome.Eur Respir J. 2001; 17:566–569.CrossrefMedlineGoogle Scholar
  • 43. Khosroshahi HE, Sahin SC, Akyuz Y, Ede H.Long term follow-up of four patients with Keutel syndrome.Am J Med. Genet A. 2014; 164A:2849–2856. doi: 10.1002/ajmg.a.36699.CrossrefMedlineGoogle Scholar
  • 44. Cranenburg EC, VAN Spaendonck-Zwarts KY, Bonafe L, Mittaz Crettol L, Rödiger LA, Dikkers FG, VAN Essen AJ, Superti-Furga A, Alexandrakis E, Vermeer C, Schurgers LJ, Laverman GD.Circulating matrix γ-carboxyglutamate protein (MGP) species are refractory to vitamin K treatment in a new case of Keutel syndrome.J Thromb Haemost. 2011; 9:1225–1235. doi: 10.1111/j.1538-7836.2011.04263.x.CrossrefMedlineGoogle Scholar
  • 45. Agatston AS, Janowitz WR, Hildner FJ, Zusmer NR, Viamonte M, Detrano R.Quantification of coronary artery calcium using ultrafast computed tomography.J Am Coll Cardiol. 1990; 15:827–832.CrossrefMedlineGoogle Scholar
  • 46. Davila JA, Johnson CD, Behrenbeck TR, Hoskin TL, Harmsen WS.Assessment of cardiovascular risk status at CT colonography.Radiology. 2006; 240:110–115. doi: 10.1148/radiol.2401050948.CrossrefMedlineGoogle Scholar
  • 47. Dalmeijer GW, van der Schouw YT, Magdeleyns E, Ahmed N, Vermeer C, Beulens JW.The effect of menaquinone-7 supplementation on circulating species of matrix Gla protein.Atherosclerosis. 2012; 225:397–402. doi: 10.1016/j.atherosclerosis.2012.09.019.CrossrefMedlineGoogle Scholar

Novelty and Significance

What Is New?

  • To our knowledge, there is no published study that finds a significant association between any matrix Gla-protein species and pulse wave velocity as a surrogate marker of arterial stiffness.

What Is Relevant?

  • High levels of desphospho-uncarboxylated matrix gla-protein are associated with pulse wave velocity independently of common cardiovascular risk factors in the general adult population. There is ongoing research to prove a beneficial effect of vitamin K in reducing vascular calcifications; our study is a positive preliminary result. Randomized clinical trials are needed to further explore whether vitamin K supplementation reduces cardiovascular risk.


In a general adult population, high levels of inactive matrix Gla-protein are positively associated with pulse wave velocity, the gold standard in assessing stiffening of the aorta.


eLetters should relate to an article recently published in the journal and are not a forum for providing unpublished data. Comments are reviewed for appropriate use of tone and language. Comments are not peer-reviewed. Acceptable comments are posted to the journal website only. Comments are not published in an issue and are not indexed in PubMed. Comments should be no longer than 500 words and will only be posted online. References are limited to 10. Authors of the article cited in the comment will be invited to reply, as appropriate.

Comments and feedback on AHA/ASA Scientific Statements and Guidelines should be directed to the AHA/ASA Manuscript Oversight Committee via its Correspondence page.