Natriuretic peptides (NPs) may regulate adipocyte metabolism including adiponectin. Infusion of atrial natriuretic peptide (ANP) increases plasma adiponectin in patients with heart failure. However, this relation has not been examined in a clinical setting or in myocardial infarction (MI). Accordingly, we investigated the interplay between proANP and adiponectin and the prognostic implications in patients with MI. We prospectively included 680 patients with ST-segment elevation myocardial infarction (STEMI) treated with primary percutaneous coronary intervention from September 2006 to December 2008. Blood samples were drawn immediately before percutaneous coronary intervention. Additionally, we included 40 patients with 4 obtained blood samples during STEMI. Adiponectin and proANP were measured in all plasma samples. All patients were followed for 5 years. End points were all-cause mortality (n = 137) and the combined end point (n = 170) of major adverse cardiovascular events (MACEs). Plasma adiponectin and proANP were strongly associated at admission (r = 0.34, p <0.001). In patients with increasing proANP during STEMI, adiponectin also increased (0.5 ± 0.3 vs −0.1 ± 0.1 mg/L, p = 0.026). During follow-up, patients with higher adiponectin at admission had increased risk of all-cause mortality and MACE (both, p <0.001). After adjustment for confounding risk factors by Cox regression analysis, adiponectin remained an independent predictor of all-cause mortality and MACE: hazard ratio 1.31 (95% confidence interval 1.07 to 1.60; p = 0.009) and 1.31 (95% confidence interval 1.09 to 1.57; p = 0.004), respectively, for each SD increase. However, the association vanished when proANP was included in the analysis. In conclusion, adiponectin is associated with an increased risk of all-cause mortality and MACE. However, concomitantly elevated proANP levels appear to confound the association between adiponectin and worsened outcome.
Adiponectin is an adipose-specific, insulin-sensitizing hormone with anti-inflammatory and anti-atherogenic effects that appears to protect against arteriosclerosis and myocardial damage. Despite these apparent beneficial cardiovascular effects, epidemiologic data have reported high adiponectin levels to be associated with impaired outcome. Recently, the relation with natriuretic peptides (NP) has emerged as a possible explanation for this “adiponectin paradox” as NP has been suggested to regulate adipocyte metabolism by inducing lipolysis, lipid oxidation, and adipocyte browning. In agreement, high levels of NP associate with a favorable adipose profile (reduced visceral and liver fat), and in both in healthy humans and patients with heart failure (HF), adiponectin and NP are strongly correlated. Interestingly, the addition of atrial natriuretic peptide (ANP) to cultured adipocytes significantly increased adiponectin expression and secretion, and in patients with HF, infusion of ANP increased plasma adiponectin levels compared with placebo infusion. Together, these findings indicate that NP may regulate plasma adiponectin levels.
However, whether this relation exists in patients with acute myocardial infarction (AMI) is unknown. Accordingly, we investigated the interplay between adiponectin and NP and the prognostic implications in a large cohort of patients with ST-segment elevation myocardial infarction (STEMI).
Methods
We prospectively included 680 patients with STEMI treated with primary percutaneous coronary intervention (PCI) from September 2006 to December 2008 (group A). Inclusion criteria were patients admitted because of a suspected STEMI with the presence of chest pain for >30 minutes and <12 hours and persistent ST-segment elevation ≥2 mm in at least 2 contiguous precordial electrocardiographic leads or ≥1 mm in at least 2 contiguous limb electrocardiographic leads or a newly developed left bundle branch block. Exclusion criteria were a nonsignificant troponin I (TnI) increase (≤0.5 μg/L), and PCI was not performed (no stenosis on coronary artery angiography or coronary artery bypass surgery was elected instead). In total, 735 patients were included; however, 55 were excluded because of missing values of proANP. To investigate longitudinal changes in proANP and adiponectin during admission for STEMI, we also included 40 patients with STEMI with 4 blood samples obtained during the admission and measurement of both proANP and adiponectin included from February 2008 to March 2011 (group B). Inclusion and exclusion criteria were similar; although to obtain a more homogeneous study population, only patients without a history of MI and an left anterior descending artery occlusion treated with a successful PCI were included. All patients (both groups A and B) were included at the Gentofte University Hospital, Denmark, a high-volume PCI center with present on-site cardiac surgery performing >1,500 PCI procedures a year. Subsequent medical treatment included daily aspirin 75 mg, clopidogrel 75 mg (for 12 months), lipid-lowering drugs (statins), and β-receptor antagonists.
We defined hypertension, hypercholesterolemia, and diabetes mellitus as use of blood pressure–lowering, cholesterol-lowering, and glucose-lowering drugs, respectively. Multivessel disease was defined as 2- or 3-vessel disease and complex lesions as type C lesions. End points were all-cause mortality and major adverse cardiovascular events (MACEs) defined as cardiovascular mortality or re-admission because of a new AMI, ischemic stroke, or symptomatic HF. AMI was defined as admission to hospital because of chest pain and a significant increase in TnI. Ischemic stroke was defined as admission to hospital because of neurologic symptoms (focal neurologic deficit) and verified by cerebral computed tomographic scan. Symptomatic HF was defined as admission to hospital with clinical signs of HF (dyspnea, fatigue, and radiographic evidence of pulmonary edema/stasis).
In group A, all 680 patients were followed for 5 years, and follow-up was 100% complete (n = 680). Data on mortality were obtained through the civil registration system and the National Causes of Death Registry, which provides information from physicians on causes of death according to the International Classification of Diseases, Tenth Revision . Follow-up data on re-MI, ischemic stroke, and admission with HF were obtained using data from the highly validated National Patient Registry and provided by ICD-10 codes. Importantly, all events found using these registries were carefully validated using medical records (e.g., laboratory tests, operative reports, hospital summaries, etc.), and possibly misclassified events excluded.
The study was approved by the local scientific ethical committee and the Danish Data Protection Agency and complied with the second Declaration of Helsinki. Informed content was obtained from all participants.
Blood samples were drawn from the femoral sheath before the PCI procedure. In group B (n = 40), blood samples were obtained at 4 times (before PCI, after PCI, and on the next 2 days). Blood samples were centrifuged within 30 minutes, and plasma was stored in nunc cryotubes at −80°C until analysis in a blinded fashion in a dedicated core laboratory. Plasma adiponectin was determined by a validated in-house time-resolved immunofluorometric assay as previously described. All samples were analyzed in duplicate, with a detection limit of 1.5 μg/L and intra- and interassay coefficients of variation <5% and <7%, respectively. We used the precursor proANP as measurement of NP as it is more stable in plasma and generally accepted to reflect ANP activity. Midregional fragment of proANP (MR-proANP) was measured using an automated processing assay (automated platform; Kryptor Compact Plus, Brahms, Germany) capable of detecting all fragments of proANP in the circulation, indifferent of endeoproteolytic cleavage. Assay performance has previously been published. C-reactive protein (CRP), TnI, and creatinine were assayed by routine laboratory methods (different TnI assays were used for the 2 populations, making comparisons of values impossible). TnI was measured at baseline and again after 6 and 12 hours. Estimated glomerular filtration rate was calculated on the basis of serum creatinine, age, and gender using the Modification of Diet in Renal Disease formula.
Plasma adiponectin, proANP, CRP, and TnI concentrations were positively skewed, and therefore, logarithmically transformed using the base logarithm of 2 before further analysis. Baseline associations between variables were evaluated using linear regression models and validated with plots of residuals, fitted values, and leverage. Linearity was tested using cubic splines. The prognostic value of adiponectin was assessed by Cox proportional hazards regression models in both univariable and multivariable analyses. Evaluation of first-order interactions was made after selecting the final model, adjusting for multiple testing by the Bonferroni method. Deviation from linearity was assessed by simultaneous assessment of linear and quadratic effects. Mis-specification of the functional form of the covariates and the assumption of proportional hazards were evaluated by plots of the cumulative martingale residuals. Statistical calculations were performed using the SAS Statistical Software (SAS for Windows, release 9.2; SAS Institute Inc., Cary, North Carolina).
Results
The plasma level of adiponectin at admission (geometric mean fifth to ninety-fifth percentile) was 7.3 mg/L (2.8 to 18.4 mg/L) and 174 pmol/L (45 to 676 pmol/L) for MR-proANP. During the 5 years of follow-up, all-cause mortality was 20% (n = 137) including 50% (n = 68) due to CV death. A total of 25% (n = 170) reached the combined end point of MACE. Table 1 describes baseline characteristics of the patients with STEMI. Baseline associations between adiponectin and covariates are listed in Table 2 . In a multivariable analysis, adiponectin associated positively with age, female gender, high-density lipoprotein (HDL), and proANP and inversely with body mass index (BMI) and triglyceride (all, p <0.001). In fact, in the multivariable analysis, proANP was the variable explaining most of the variation in plasma adiponectin, followed by HDL, male gender, BMI, triglyceride, and age, all variables well known for their association with adiponectin.
| Variable | Group A (n=680) | Group B (n=40) |
|---|---|---|
| Age (years) ∗ | 63±12 | 59±13 |
| Men | 503 (74%) | 35 (88%) |
| Hypertension | 231 (34%) | 5 (13%) |
| Diabetes mellitus | 61 (9%) | 3 (8%) |
| Current smoker | 347 (51%) | 18 (45%) |
| Hypercholesterolemia | 129 (19%) | 4 (10%) |
| Previous myocardial infarction | 41 (6%) | 0 (0%) |
| Body mass index (kg/m 2 ) ∗ | 26.7±4.5 | 27.0±3.6 |
| Blood glucose (mmol/L) † | 8.3 (7.0-9.9) | 7.9 (6.5-10.5) |
| Total Cholesterol (mmol/L/mg/dl) ∗ | 4.8±1.1/186±43 | 5.7±1.0/220±39 |
| High density lipoprotein (mmol/L/mg/dl) ∗ | 1.3±0.4/50±15 | 1.7±0.4/66±15 |
| Low density lipoprotein (mmol/L/mg/dl) ∗ | 2.9±1.0/112±39 | 3.8±0.9/147±35 |
| Triglyceride (mmol/L/mg/dl) † | 1.0 (0.7-1.6)/86 (62-142) | 1.0 (0.6-1.9)/86 (53-168) |
| Peak troponin I (μg/L) † ‡ | 95 (28-248) | 44 (17-100) |
| C reactive protein (mg/L) † | 3 (1-9) | 4 (2-6) |
| Serum creatinine (μmol/L ) † | 92 (77-109) | 73 (615-81) |
| Estimated glomerular filtration rate (mL/min) † | 73±24 | 100±28 |
| Symptom-to-balloon time (min) † | 196 (130-328) | 170 (136-298) |
| Complex lesion | 333 (49%) | 15 (37%) |
| Multivessel coronary artery disease | 184 (27%) | 0 (0%) |
| Left anterior descending coronary artery narrowing | 306 (45%) | 40 (100%) |
| Glycoprotein IIb/IIIa inhibitor | 170 (25%) | 13 (33%) |
| Adiponectin (mg/L) | 6.9 (5.0-10.2) | 5.6 (4.5-7.4) |
| Pro-atrial natriuretic peptide (pmol/L) † | 168 (111-267) | 138 (85-213) |
∗ Gaussian distributed as mean ± standard deviation.
† Non-Gaussian distributed as median (interquartile range).
‡ Not measured with the same assay making comparison impossible.
| Variables | Log 2 -adiponectin | ||||
|---|---|---|---|---|---|
| Univariable | Multivariable | ||||
| β | p-value | β | Explanation of variation in adiponectin | p-value | |
| Age (years) | 0.026 | <0.001 | 0.011 | 9% | <0.001 |
| Men | -0.673 | <0.001 | -0.309 | 13% | <0.001 |
| Hypertension | 0.070 | 0.29 | |||
| Diabetes mellitus | -0.393 | <0.001 | |||
| Current smoker | -0.182 | 0.003 | |||
| Hypercholesterolemia | -0.265 | <0.001 | |||
| Previous myocardial infarction | -0.195 | 0.15 | |||
| Systolic blood pressure (mmHg) | 0.001 | 0.55 | |||
| Body mass index (kg/m 2 ) | -0.061 | <0.001 | -0.028 | 11% | <0.001 |
| Blood glucose (mmol/L) | -0.133 | 0.08 | |||
| Total Cholesterol (mmol/L/mg/dl) | 0.008/0.000 | 0.94 | |||
| High density lipoprotein (mmol/L/mg/dl) | 0.888/0.023 | <0.001 | 0.493/0.013 | 22% | <0.001 |
| Low density lipoprotein (mmol/L/mg/dl) | -0.013/-0.000 | 0.48 | |||
| Log2-triglyceride (mmol/L/mg/dl) | -0.306/-0.003 | <0.001 | -0.137/-0.002 | 10% | <0.001 |
| Log2-pro-atrial natriuretic peptide (pmol/L) | 0.307 | <0.001 | 0.224 | 27% | <0.001 |
| Log2-peak troponin I (μg/L) | -0.001 | 0.93 | |||
| Log2-C-reactive protein (mg/L) | 0.009 | 0.60 | |||
| Estimated glomerular filtration rate (mL/min) | -0.004 | <0.001 | |||
| Symptom-to-balloon time (min) | -0.001 | 0.99 | |||
| Complex lesion | -0.002 | 0.97 | |||
| Multivessel coronary artery disease | -0.041 | 0.59 | |||
| Left anterior descending coronary artery narrowing | 0.005 | 0.94 | |||
| Glycoprotein IIb/IIIa inhibitor | -0.087 | 0.22 | |||
Figure 1 shows adiponectin levels for increasing proANP quintiles. As seen, adiponectin increased in a linear fashion for each increasing quintile of proANP, indicating a dose-response relation. We then tested if the association between proANP and adiponectin was linear using cubic splines. The linear association was highly significant (p linear <0.001). In contrast, a possible nonlinear association was clearly nonsignificant (p nonlinear = 0.55) confirming the dose-response relation.
For a more detailed investigation of the relation between adiponectin and proANP, we also measured adiponectin and proANP in a group of patients with STEMI (n = 40; group B) with 4 consecutive blood samples obtained during the admission. Overall, during admission for STEMI, proANP decreased (from 133 ± 14 to 96 ± 9 pmol/L, p <0.001), whereas adiponectin remained unchanged (from 5.7 ± 0.4 to 5.8 ± 0.3 mg/L, p = 0.51; Figure 2 ). Similar to patients in group A, adiponectin and proANP were associated at admission (β = 0.311 [0.105 to 0.517], p = 0.003). Although mean proANP levels decreased during the admission, proANP increased in 6 patients (15%). In these 6 patients, adiponectin increased significantly compared with those with unchanged/decreasing proANP (0.5 ± 0.3 vs −0.1 ± 0.1 mg/L, p = 0.026). Even when adjusted for all variables significantly associated with adiponectin (age, gender, diabetes, BMI, triglyceride, and HDL) together with admission levels of adiponectin and proANP, increasing proANP associated independently with increasing adiponectin (β = 0.76 [0.25 to 1.26], p = 0.003).
