OUP user menu

Parathyroid hormone level is associated with mortality and cardiovascular events in patients undergoing coronary angiography

Stefan Pilz, Andreas Tomaschitz, Christiane Drechsler, Eberhard Ritz, Bernhard O. Boehm, Tanja B. Grammer, Winfried März
DOI: http://dx.doi.org/10.1093/eurheartj/ehq109 1591-1598 First published online: 3 May 2010

Abstract

Aims Elevated parathyroid hormone (PTH) levels have been associated with increased cardiovascular risk in the general population. We aimed to elucidate whether PTH levels are associated with mortality and fatal cardiovascular events in patients referred for coronary angiography.

Methods and results Intact PTH was measured in 3232 Caucasian patients from the LUdwigshafen RIsk and Cardiovascular Health (LURIC) study, who underwent coronary angiography at baseline (1997–2000). During a median follow-up time of 7.7 years, 742 patients died including 467 deaths due to cardiovascular causes. Unadjusted Cox proportional hazard ratios (HRs) (with 95% confidence intervals) in the fourth when compared to the first PTH quartile were 2.13 (1.75–2.60) for all-cause and 2.47 (1.92–3.17) for cardiovascular mortality. After adjustments for common cardiovascular risk factors, these HRs remained significant with 1.71 (1.39–2.10) for all-cause and 2.02 (1.55–2.63) for cardiovascular mortality. Among specific cardiovascular events we observed a particularly strong association of PTH with sudden cardiac death (SCD). The adjusted HR for SCD in the first vs. the fourth PTH quartile was 2.68 (1.71–4.22).

Conclusion Our results among patients undergoing coronary angiography show that PTH levels are an independent risk factor for mortality and cardiovascular events warranting further studies to evaluate whether PTH modifying treatments reduce cardiovascular risk.

  • PTH
  • Vitamin D
  • Cardiovascular disease
  • Sudden cardiac death
  • Mortality
  • Prospective study

Introduction

Parathyroid hormone (PTH), which is crucial for the maintenance of calcium homeostasis, has been associated with increased cardiovascular risk.1 Parathyroid hormone is secreted by the parathyroid glands in response to hypocalcaemia which is detected by the calcium sensing receptor (CaSR).2 Classic PTH effects on bone and kidney are important for the control of calcium homeostasis, but PTH receptors are also expressed in the vessel walls and the myocardium suggesting direct effects on the cardiovascular system.1 In this context, PTH levels have been associated with hypertension, myocardial dysfunction, and vascular diseases.16 Primary hyperparathyroidism, characterized by inadequately high PTH levels with subsequent hypercalcaemia, has been associated with increased cardiovascular risk and mortality, which could be significantly reduced by parathyroidectomy in most but not all studies in this field.1,2,7 Secondary hyperparathyroidism, which is frequently observed in patients with impaired kidney function, is also associated with increased mortality and PTH modifying therapies have been shown to improve the clinical outcome of patients with renal failure.2,8 Only a few studies addressed the clinical significance of PTH levels for mortality and cardiovascular diseases in persons without significant renal disease.2,7 Currently available studies in this field have shown an association of PTH levels and mortality in the elderly.2,915 Importantly, a recent study among a community-based cohort of 958 elderly men has shown that PTH levels within the normal range are predictive for fatal cardiovascular events.13 Whether this applies for both genders and for patients at higher cardiovascular risk remains largely unknown. Given that PTH modifying therapies such as vitamin D and calcium supplementation can be easily and relatively safely performed, there exists a great public health interest to elucidate whether PTH is a promising target for the treatment to reduce cardiovascular risk.1,2 Hence, we aimed to prospectively evaluate whether PTH levels are a risk factor for mortality and fatal cardiovascular events [e.g. sudden cardiac death (SCD)] in a large cohort of patients referred for coronary angiography.

Methods

Participants and study design

The LURIC study, a prospective cohort study of patients referred to coronary angiography, was designed to investigate environmental and genetic risk factors for cardiovascular diseases.16 The baseline examination was performed between July 1997 and January 2000 at a tertiary care centre in south-west Germany (Herzzentrum Ludwigshafen) and included 3316 study participants. Inclusion criteria were the availability of a coronary angiogram, clinical stability with the exception of acute coronary syndromes (ACSs), and Caucasian origin, in order to limit genetic heterogeneity. Indications for coronary angiography were commonly chest pain or non-invasive tests in which myocardial ischaemia was suspected. Patients with any acute illness other than ACS, with a history of malignancy within the past 5 years and with any predominant non-cardiac disease were excluded from the study. Informed written consent was obtained from all study participants and approval for the study was obtained from the ethics committee at the ‘Ärztekammer Rheinland-Pfalz’ (Mainz, Germany). The LURIC study complies with the Declaration of Helsinki.

Baseline examination

Detailed descriptions of the baseline examination in LURIC have been published previously.16 Angiographic coronary artery disease (CAD) was diagnosed in patients with at least one stenosis ≥50% of at least one out of 15 coronary segments, using the maximal luminal narrowing estimated by visual analysis. Diabetes mellitus was diagnosed if the fasting glucose was >7.0 mmol/L or the 2 h value in an oral glucose tolerance test was >11.1 mmol/L and in patients already receiving antidiabetic medication. Arterial hypertension was diagnosed if the mean systolic and diastolic blood pressures out of five measurements exceeded 140 and/or 90 mmHg or if patients were already on antihypertensive treatment.

Biochemical analyses

Routine laboratory measurements were performed as previously described.16 In brief, venous blood sampling was performed in the morning before coronary angiography and routine laboratory parameters were immediately determined, whereas remaining blood samples were snap frozen for further determinations and stored at −80°C until analysis. Intact PTH was determined in serum by ElectroChemiLuminescence Immunoassay (ECLIA) on an Elecsys 2010 (Roche Diagnostics, Mannheim, Germany), with a normal range of 15–65 pg/mL and an inter-assay coefficient of variation of 5.7–6.3%. Serum concentrations of 25(OH)D were measured by a radioimmunoassay (DiaSorin Antony, France; Stillwater, USA) with an intra- and inter-assay coefficient of variation of 8.6 and 9.2%, respectively.17 Levels of 1,25-dihydroxyvitamin D were also measured by radioimmunoassay (Nichols Institute Diagnostika GmbH, Bad Nauheim, Germany) on a multicrystal counter (Berthold LB2014, DiaSorin, SA, USA). N-terminal pro-B-type natriuretic peptide (NT-pro-BNP) was determined by ElectroChemiLuminescence (ECL) on an Elecsys 2010 (Roche Diagnostics, Mannheim, Germany). C-reactive protein was measured by immunonephelometry (N High Sensitivity CRP, Dade Behring, Marburg, Germany). Plasma aldosterone was determined by radioimmunoassay (Active aldosterone, Diagnostic Systems Laboratories, Sinsheim, Germany) and glomerular filtration rate (GFR) was calculated according to the abbreviated MDRD study equation.18

Follow-up

Information on vital status was obtained from local person registries. Hospital records and death certificates were reviewed to classify the causes of death. Cardiovascular deaths included SCD, fatal myocardial infarction, deaths due to heart failure, deaths after intervention to treat CAD, stroke, and other deaths due to heart disease. Sudden cardiac death was defined as a sudden unexpected death either within 1 h of symptom onset or within 24 h of having been observed alive and symptom free.19,20 Persons whose sudden death was most likely attributable to a non-cardiac disease and patients who suffered from any predominant non-cardiac and terminal disease (e.g. cancer) so that their death was not unexpected were not classified as having died due to SCD. Three experienced clinicians who were blinded to PTH values and baseline characteristics of the study probands classified the causes of death.

Statistics

Parathyroid hormone quartiles were calculated according to the values of the entire study cohort. All skewed continuous parameters were logarithmically transformed before use in parametric procedures. Baseline characteristics were stratified by PTH quartiles. Depending on their distribution, continuous parameters are either presented as means ± standard deviation (normally distributed variables) or as medians with interquartile range (skewed variables). Categorical data are shown as proportions. Comparisons across PTH quartiles were calculated by analysis of variance (ANOVA) with P for trend for continuous parameters and by χ2 test with P for liner-by-linear test for categorical variables. For all-cause and cardiovascular mortality, Kaplan–Meier curves followed by a log-rank test were graphed to show the differences in event-free survival between PTH quartiles. Hazard ratios (HRs) with 95% confidence intervals (95% CI) for all-cause and cardiovascular mortality were calculated with Cox proportional hazard models using the first PTH quartile as the reference. We present data from these analyses of a crude (unadjusted) model, of an age- and sex-adjusted model (Model 1) and of a model adjusted for common cardiovascular risk factors (Model 2). The latter model includes variables for BMI (kg/m2), haemoglobin A1c (HbA1c) (%), systolic blood pressure (mmHg), GFR (mL/min/1.73 m2), LDL- and HDL-cholesterol (mg/dL), triglycerides (mg/dL), C-reactive protein (mg/L), ex- and active smokers (yes/no), and number of diseased vessels (0–3 vessels). In addition to the covariates of Model 2, we adjusted for the use of medication (ACE-inhibitors, beta-blockers, diuretics, aspirin/other antiplatelet agent, and statins), ventricular dysfunction (NT-pro-BNP) (ng/mL), markers of malnutrition (haemoglobin in g/dL and albumin in g/dL), serum calcium (mmol/L), or 25(OH)D (ng/mL). A further model (Model 3) included all covariates of Model 2 plus various parameters of mineral metabolism (serum calcium, serum phosphate, 25[OH]D, and use of diuretics). We also calculated HRs for patients within the reference range of PTH (≤65 pg/mL) to test whether our findings also apply for individuals without hyperparathyroidism. In addition, we tested for interactions by adding product terms to the multivariable adjusted models (Model 2) in order to examine whether the association of PTH quartiles with mortality differs according to sex and the presence or absence of ACS, CAD, significant renal failure (defined as GFR ≤ 60 mL/min/1.73 m2), or vitamin D deficiency (defined as 25[OH]D below 20 ng/mL). Finally, we calculated the C statistic (equivalent to the area under the receiver operating characteristic curve) to assess the discriminative power of the multivariate adjusted model (Model 2) with and without PTH. A P-value < 0.05 was considered statistically significant and all statistical tests were two-sided. Data were analysed using SPSS 15.0 statistical package (SPSS Inc., Chicago, IL, USA).

Results

Parathyroid hormone levels were available in 3232 study participants and elevated PTH levels above 65 pg/mL which are indicative for hyperparathyroidism were observed in 174 patients (5.4%). Only three patients with hyperparathyroidism displayed hypercalcaemia, defined as serum calcium levels above 2.65 mmol/L, indicative for primary hyperparathyroidism. Clinical and laboratory baseline characteristics according to PTH quartiles are shown in Table 1. Coronary artery disease was ruled out in approximately one-third (31.9%) of the study population. Age, BMI, blood pressure, C-reactive protein, aldosterone, and NT-pro-BNP were significantly increased and triglycerides, GFR, haemoglobin, albumin, 25(OH)D, 1,25 (OH)2D, serum calcium, and serum phosphate were significantly decreased in higher PTH quartiles. Proportion of ex- and active smokers, patients with ACS, and use of beta-blockers were significantly reduced and females, higher NYHA classes, frequency of atrial fibrillation and ACS, and the use of ACE-inhibitors and diuretics were significantly increased in patients with higher PTH quartiles.

View this table:
Table 1

Baseline characteristics stratified by parathyroid hormone quartiles

PTH quartilesP-value
1st quartile2nd quartile3rd quartile4th quartile
PTH (pg/mL)<2323–2930–40>40
Numbers885742847758
Age (years)59.9 (52.4–67.3)62.5 (55.8–69.3)64.7 (58.2–71.0)67.8 (60.2–72.8)<0.001
Female (%)26.129.831.834.7<0.001
Body mass index (kg/m2)26.6 (24.7–29.2)27.1 (24.9–29.4)27.2 (24.5–29.9)27.4 (24.8–30.1)0.002
Diabetes mellitus (%)30.531.431.135.20.067
HbA1c (%)6.0 (5.6–6.6)6.0 (5.5–6.5)6.0 (5.6–6.6)6.1 (5.6–6.8)0.088
Arterial hypertension (%)67.970.673.879.4<0.001
Blood pressure (mmHg)
 Systolic138 ± 23141 ± 23142 ± 23145 ± 24<0.001
 Diastolic79 ± 1181 ± 1182 ± 1182 ± 12<0.001
Ex- and active smokers (%)69.063.762.859.2<0.001
Blood lipids (mg/dL)
 LDL-cholesterol114 (95–137)115 (95–138)113 (93–137)115 (93–138)0.360
 HDL-cholesterol37 (31–44)37 (32–45)38 (32–45)37 (30–45)0.970
 Trigylycerides154 (112–217)141 (108–197)145 (105–191)141 (108–196)<0.001
GFR (mL/min per 1.73 m2)84.5 (73.2–95.0)81.9 (71.6–93.5)81.1 (70.7–91.7)76.1 (60.6–89.0)<0.001
C-reactive protein (mg/L)3.3 (1.2–8.1)3.0 (1.2–7.9)3.2 (1.3–8.5)4.1 (1.5–9.4)0.005
Aldosterone (pg/mL)73 (45–117)75 (46–119)79 (46–121)86 (56–143)<0.001
Coronary artery disease (%)67.769.067.268.40.880
Number of vessels disease (%)
 No stenosis32.331.032.831.60.609
 Single-vessel disease19.519.518.718.2
 Two-vessel disease17.921.418.518.8
 Three-vessel disease30.328.130.031.4
Acute coronary syndrome (%)33.631.430.728.10.019
History of MI (%)42.141.138.541.80.587
NT-pro-BNP (ng/mL)213 (87–559)257 (97–733)312 (111–847)501 (177–1888)<0.001
NYHA class (%)
 NYHA 159.156.249.642.2<0.001
 NYHA 225.926.332.532.2
 NYHA 312.814.815.221.4
 NYHA 42.32.72.74.2
Atrial fibrillation (%)7.48.415.018.1<0.001
Medication use (%)
 ACE-inhibitor49.653.551.758.60.001
 Beta-blocker65.664.365.557.10.002
 Aspirin/other platelet agent71.074.572.066.90.058
 Statin49.048.145.345.10.061
 Diuretics21.424.326.243.8<0.001
Haemoglobin (g/dL)14.0 ± 1.513.9 ± 1.413.8 ± 1.513.6 ± 1.5<0.001
Albumin (g/dL)4.4 (4.1–4.8)4.3 (4.0–4.7)4.3 (4.0–4.8)4.3 (4.0–4.7)0.030
25(OH)D (ng/mL)19.0 (12.2–26.1)16.9 (11.5–23.6)14.8 (9.8–21.7)12.2 (7.7–18.4)<0.001
1,25(OH)2 D (pg/mL)33.4 (25.7–42.9)34.5 (26.4–44.0)33.4 (25.5–42.6)31.7 (23.5–42.1)0.001
Serum calcium (mmol/L)2.35 (2.28–2.41)2.33 (2.26–2.40)2.32 (2.25–2.38)2.31 (2.24–2.38)<0.001
Serum phosphate (mg/dL)3.6 (3.2–4.0)3.5 (3.2–3.9)3.5 (3.1–3.8)3.4 (3.0–3.8)<0.001
  • Continuous data are presented as means ± standard deviation and as medians with interquartile range and categorical data are shown as percentages. ANOVA with P for trend and χ2 test were used.

  • GFR, glomerular filtration rate; MI, myocardial infarction; NT-pro-BNP, N-terminal-pro-B-type natriuretic peptide; 25(OH)D, 25-hydroxyvitamin D; 1,25(OH)2D, 1,25-dihydroxyvitamin D.

Parathyroid hormone, mortality, and fatal cardiovascular events

Eighteen patients were lost during a median follow-up time of 7.7 years (interquartile range: 7.2–8.5 years) and in 24 study participants we did not obtain sufficient information to classify their causes of death. These latter patients were included in the analyses for all-cause mortality but excluded from analyses for cardiovascular mortality and specific cardiovascular events. Among 3232 patients with available PTH levels, 742 died during follow-up, of whom 467 died due to cardiovascular causes. Among specific cardiovascular events we recorded 187 SCDs, 84 fatal myocardial infarctions, 112 deaths due to heart failure, and 84 deaths due to other cardiovascular causes. Kaplan–Meier curves followed by a log-rank test showed that all-cause and cardiovascular mortality significantly increased in the highest two PTH quartiles (P < 0.001 for both; Figure 1A and B). In detail, there was a J-shaped association with the lowest mortality risk in the second PTH quartile and a significantly increased mortality risk in the third PTH quartile and even more pronounced in the fourth PTH quartile.

Figure 1

(A) Kaplan–Meier curves for all-cause mortality according to PTH quartiles. (B) Kaplan–Meier curve for cardiovascular mortality according to PTH quartiles.

Unadjusted HRs (with 95% CI) for all-cause and cardiovascular mortality in the fourth (highest) PTH quartile when compared with the first (lowest) PTH quartile were 2.13 (1.75–2.60) and 2.47 (1.92–3.17), respectively (Table 2). After adjustments for common cardiovascular risk factors, these HRs were attenuated to 1.71 (1.39–2.10) for all-cause mortality and 2.02 (1.55–2.63) for cardiovascular mortality. These HRs remained significant even after further adjustments for the use of medication, NT-pro-BNP, albumin and haemoglobin, serum calcium, 25(OH)D, or a combination of various parameters of mineral metabolism (Table 2). Adjustments for albumin-corrected serum calcium or aldosterone did also not materially change our results (data not shown). For all-cause mortality, the C statistic for Model 2 was 0.759 (95% CI: 0.739–0.778) without PTH and 0.765 (0.746–0.785) with PTH. For cardiovascular mortality, the C statistic was 0.748 (0.724–0.771) without PTH and 0.758 (0.734–0.782) with PTH.

View this table:
Table 2

Hazard ratios with 95% confidence intervals for all-cause and cardiovascular mortality according to parathyroid hormone quartiles

PTH quartiles1st quartile2nd quartile3rd quartile4th quartile
Range of values (pg/mL)<2323–2930–40>40
All-cause mortality
 Study participants at risk877738843756
 Number of deaths158 (18.0%)125 (16.9%)200 (23.7%)259 (34.3%)
 Mean survival in years (±SE)8.46 ± 0.088.61 ± 0.088.13 ± 0.097.43 ± 0.12
 Crude model1.0 reference0.92 (0.73–1.17)1.36 (1.11–1.68)2.13 (1.75–2.60)
 Model 1a1.0 reference0.82 (0.65–1.04)1.07 (0.87–1.32)1.54 (1.26–1.88)
 Model 2b1.0 reference0.94 (0.74–1.19)1.24 (1.00–1.54)1.71 (1.39–2.10)
 Model 2 plus medication usec1.0 reference0.95 (0.75–1.21)1.27 (1.03–1.58)1.48 (1.20–1.83)
 Model 2 plus NT-pro-BNP1.0 reference0.92 (0.73–1.16)1.21 (0.97–1.50)1.51 (1.22–1.86)
 Model 2 plus albumin and haemoglobin1.0 reference1.01 (0.78–1.30)1.23 (0.98–1.56)1.76 (1.40–2.20)
 Model 2 plus serum calcium1.0 reference0.95 (0.75–1.21)1.26 (1.02–1.57)1.74 (1.41–2.15)
 Model 2 plus 25-hydroxyvitamin D1.0 reference0.90 (0.71–1.14)1.14 (0.92–1.41)1.38 (1.12–1.71)
 Model 3d1.0 reference0.93 (0.73–1.18)1.19 (0.96–1.48)1.36 (1.09–1.69)
Cardiovascular mortality
 Study participants at risk873734835748
 Number of deaths94 (10.8%)70 (9.5%)124 (14.9%)179 (23.9%)
 Mean survival in years (±SE)8.86 ± 0.078.98 ± 0.068.54 ± 0.087.93 ± 0.11
 Crude model1.0 reference0.87 (0.64–1.19)1.42 (1.09–1.86)2.47 (1.92–3.17)
 Model 1a1.0 reference0.78 (0.57–1.07)1.14 (0.87–1.49)1.81 (1.41–2.34)
 Model 2b1.0 reference0.91 (0.67–1.25)1.35 (1.03–1.78)2.02 (1.55–2.63)
 Model 2 plus medication usec1.0 reference0.93 (0.68–1.28)1.40 (1.06–1.84)1.76 (1.35–2.29)
 Model 2 plus NT-pro-BNP1.0 reference0.89 (0.65–1.21)1.31 (0.99–1.73)1.77 (1.35–2.31)
 Model 2 plus albumin and haemoglobin1.0 reference0.98 (0.70–1.37)1.36 (1.01–1.84)2.09 (1.57–2.78)
 Model 2 plus serum calcium1.0 reference0.93 (0.68–1.28)1.39 (1.06–1.84)2.10 (1.61–2.74)
 Model 2 plus 25-hydroxyvitamin D1.0 reference0.88 (0.64–1.20)1.26 (0.95–1.66)1.66 (1.27–2.17)
 Model 3d1.0 reference0.91 (0.66–1.24)1.32 (1.00–1.75)1.65 (1.25–2.17)
  • aAdjusted for age and sex.

  • bAdditionally adjusted for body mass index, ex- and active smokers, HbA1c, systolic blood pressure, glomerular filtration rate (GFR), LDL- and HDL-cholesterol, triglycerides, C-reactive protein, number of diseased vessels.

  • cUse of ACE-inhibitors, beta-blockers, diuretics, aspirin/other antiplatelet agent, and statins.

  • dModel 3 includes all covariates of Model 2 plus serum calcium, serum phosphate, 25-hydroxyvitamin D, and use of diuretics.

After exclusion of patients with hyperparathyroidism defined as PTH levels above 65 pg/mL, the HRs in the fourth vs. the first PTH quartile remained significant for all-cause [1.49 (1.19–1.86)] and for cardiovascular mortality [1.74 (1.31–2.31)] after adjustments for common cardiovascular risk factors (same adjustments as in Model 2 in Table 2).

In analyses for all-cause mortality, there were no significant interactions of PTH quartiles with ACS (P = 0.408), CAD (P = 0.597), significant renal disease (P = 0.259), and vitamin D deficiency (P = 0.289). In subgroup analyses, adjusted HRs for all-cause mortality (according to Model 2 in Table 2) in the fourth vs. the first PTH quartile were 1.87 (1.30–2.67) for patients with ACS (n = 1003), 1.61 (1.25–2.08) for patients without ACS (n = 2229), and 1.69 (1.25–2.29) for patients with stable CAD (no ACS but CAD; n = 1248). For further adjusted HRs of the respective subgroup analyses see Figure 2A. In analyses for cardiovascular mortality, there were also no significant interactions of PTH quartiles with ACS (P = 0.990), CAD (P = 0.622), significant renal disease (P = 0.304), and vitamin D deficiency (P = 0.460). In subgroup analyses, adjusted HRs for cardiovascular mortality in the fourth vs. the first PTH quartile were 2.04 (1.48–2.81) for patients with ACS, 1.91 (1.20–3.06) for patients without ACS, and 2.01 (1.39–2.90) for patients with stable CAD. For further adjusted HRs of the respective subgroup analyses see Figure 2B.

Figure 2

(A) Forest plots of Cox proportional hazard ratios with 95% CI (adjusted for cardiovascular risk factors) for all-cause mortality in the fourth vs. the first PTH quartile. Results are shown for all study participants and for subgroups stratified by the presence of coronary artery disease, acute coronary syndrome, significant renal disease (defined as GFR ≤60 mL/min/1.73 m2) and vitamin D deficiency (defined as 25[OH]D levels <20 ng/mL). (B) Forest plots of Cox proportional hazard ratios with 95% CI (adjusted for cardiovascular risk factors) for cardiovascular mortality in the fourth vs. the first PTH quartile. Results are shown for all study participants and for subgroups stratified by the presence of coronary artery disease, acute coronary syndrome, significant renal disease (defined as GFR ≤60 mL/min/1.73 m2) and vitamin D deficiency (defined as 25[OH]D levels <20 ng/mL).

In analyses of the entire study cohort for specific cardiovascular events, the HR adjusted (Model 2 in Table 2) for common cardiovascular risk factors in the fourth vs. the first PTH quartile was 2.68 (1.71–4.22) for SCD, 1.85 (1.02–3.35) for fatal myocardial infarction, and 1.94 (1.15–3.25) for deaths due to heart failure.

All of our results did not materially change, when males and females were analysed separately and there was no significant interaction by sex for the association of PTH quartiles with all-cause mortality (P = 0.842) and cardiovascular mortality (P = 0.965).

Discussion

In patients referred to coronary angiography, PTH levels were significantly associated with all-cause and cardiovascular mortality. These results remained significant after adjustments for common cardiovascular risk factors and other possible confounders including parameters of mineral metabolism. Our findings were materially unchanged after excluding patients with hyperparathyroidism and the association of PTH with mortality was not significantly different in patients with or without CAD, ACS, significant renal disease, or vitamin D deficiency. Among specific cardiovascular events, we observed a particularly strong association of PTH with SCD.

Our findings are in line with other studies which found an association of PTH and increased mortality.2,815 However, we are, to the best of our knowledge, the first to show (i) that PTH is associated with all-cause and cardiovascular mortality in a large cohort of patients referred to coronary angiography and (ii) that PTH is an independent risk factor for SCD.2,815 Given this strong association of PTH and cardiovascular risk in epidemiological studies, it could be hypothesized that PTH itself contributes to vascular and myocardial diseases. This notion is supported by the fact that vascular smooth muscle cells, endothelial cells, and cardiomyocytes are all target cells for PTH.1,21 Clinical studies showing associations of PTH levels or primary hyperparathyroidism with endothelial dysfunction, coronary heart disease, and carotid-intima media thickness suggest pro-atherosclerotic properties of PTH.1,2,57,21,22 It should, however, be acknowledged that the underlying mechanism for these pro-atherosclerotic effects of PTH remain largely unknown. An association of PTH levels with atherosclerosis was not consistently observed in all studies,1,2,22 including our study which failed to show a significant association of PTH levels and angiographic CAD (Table 1). Another proposed mechanism linking PTH to cardiovascular diseases may be arterial hypertension because PTH infusions increased blood pressure in healthy volunteers and PTH was an independent determinant of blood pressure in a population-based study.1,3,23 We also observed an association of PTH levels with elevated blood pressure in the LURIC study. After controlling for blood pressure and use of antihypertensive medication, the association of PTH with increased cardiovascular risk remained significant suggesting other mechanisms than hypertension mediating the relationship of PTH levels and fatal cardiovascular events.

Direct effects of PTH on cardiomyocytes may be mechanistically relevant for our mortality results because PTH induces myocardial hypertrophy, increases heart rate and automaticity, and may thus increase the risk of cardiac arrhythmias and fatal cardiovascular events.24,25 In line with this, McCarty et al.24 hypothesized that PTH-induced activation of phospholipase C (PLC) may increase arrhythmias by generation of inositol-1,4,5-triphosphate, which has been associated with reperfusion arrhythmias, likewise by increasing calcium release from the sarcoplasmatic reticulum. The association of PTH with atrial fibrillation and SCD in the LURIC study fits well to the concept that PTH increases the risk for arrhythmias. Apart from this, we observed an association of PTH with NT-pro-BNP and higher NYHA class suggesting a link between PTH and heart failure. This is consistent with studies showing that PTH levels are increased in patients with heart failure and are an independent predictor of hospitalization for heart failure.26,27 Assuming a causal relationship of PTH and heart failure, it was speculated that elevated PTH levels contribute to a systemic illness that accompanies heart failure and is characterized by increased oxidative stress, a pro-inflammatory state with elevated IL-6 and TNF-α levels and a catabolic state.26,28 It should, however, be noted that increases in PTH in patients with heart failure are often simply a consequence of increased urinary calcium loss due to both the use of diuretics and secondary hyperaldosteronism.28 Furthermore, there is evidence that PTH exerts inotropic effects and improves contractile performance of cardiomyocytes.1,29 Hence, the association of PTH with heart failure is a complex interplay of various harmful and beneficial PTH effects on myocardial structure and function that need to be further clarified in detail.

Given the relatively strong and independent association of PTH with increased risk of mortality and fatal cardiovascular events, we believe that further studies are needed to elucidate whether PTH modifying therapies such as vitamin D supplementation, calcium intake, or treatment with calcimimetics (CaSR agonists) improve the clinical outcome of patients at high cardiovascular risk and with elevated or high normal PTH levels. In our opinion, in particular, vitamin D supplementation is a promising therapeutic approach. Apart from lowering PTH levels, vitamin D is suggested to exert numerous beneficial effects on the cardiovascular system, prevents fractures and falls, and may reduce other diseases such as infections and cancer.17,3033 Importantly, treatment of secondary hyperparathyroidism has already been shown to improve the clinical outcome of patients with renal failure.2 We are aware that in the LURIC study, GFR decreased with higher PTH quartiles suggesting beginning (renal) secondary hyperparathyroidism, but our results remained materially unchanged after adjustment for GFR. Importantly, our data suggest that PTH is a risk factor for cardiovascular diseases in both patients with and without significant renal disease (Figure 2A and B).

A limitation of the present work is that our results apply only to carefully selected patients undergoing coronary angiography and without predominant non-cardiac disease or malignancy. Our findings may therefore not be generalizable. Furthermore, despite careful adjustments for several possible confounders we cannot rule out residual confounding. On the other hand, our statistical models may also be over-adjusted because some covariates might lie in the causal pathway of deleterious PTH effects. Another drawback of our work is that we did not measure both parathyroid hormone-related peptide (PTH-rP),21 a structural-related peptide which shares some effects with PTH, as well as fibroblast growth factor-23 (FGF-23), which predicts mortality and prevents hyperphosphataemia by suppressing PTH and 1,25-dihydroxyvitamin D.34,35

In summary, we have shown that PTH even within the normal range is associated with mortality and fatal cardiovascular events in patients undergoing coronary angiography.

Further studies are needed to elucidate the underlying mechanisms for our results and to evaluate whether PTH modifying therapies reduce cardiovascular risk.

Funding

The LURIC study was funded by grants from the Deutsche Forschungsgemeinschaft (GRK 1041 and SFB 518) and Exzellenzzentrum ‘Stoffwechselkrankheiten’ Baden-Württemberg to B.O.B. The LURIC study was also supported by unrestricted grants from Sanovi-Aventis, Roche, Dade Behring and AstraZeneca.

Conflict of interest: Roche Diagnostics provided reagents for the measurement of PTH, but did not assume any other role in the design, conduct, or interpretation of this study.

Acknowledgements

We thank the LURIC study team either temporarily or permanently involved in patient recruitment and sample and data handling and the laboratory staff at the Ludwigshafen General Hospital and the Universities of Freiburg, Ulm and Graz and the German registration offices and local public health departments for their assistance.

References

View Abstract