If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
1 These authors contributed equally to this paper (co-first authors).
Affiliations
School of Biomedical Sciences, University of Queensland, St Lucia, Queensland, AustraliaQueensland Brain Institute, University of Queensland, St Lucia, Queensland, AustraliaUniversity of Queensland Centre for Clinical Research, University of Queensland, Herston, Queensland, AustraliaDepartment of Neurology, Royal Brisbane & Women's Hospital, Herston, Queensland, Australia
1 These authors contributed equally to this paper (co-first authors).
F.J. Steyn
Footnotes
1 These authors contributed equally to this paper (co-first authors).
Affiliations
School of Biomedical Sciences, University of Queensland, St Lucia, Queensland, AustraliaUniversity of Queensland Centre for Clinical Research, University of Queensland, Herston, Queensland, Australia
School of Biomedical Sciences, University of Queensland, St Lucia, Queensland, AustraliaThe Wesley-St Andrew's Research Institute, Auchenflower, Queensland, Australia
Department of Neurology, Royal Brisbane & Women's Hospital, Herston, Queensland, AustraliaSchool of Medicine, University of Queensland, Herston, Queensland, Australia
University of Queensland Centre for Clinical Research, University of Queensland, Herston, Queensland, AustraliaDepartment of Neurology, Royal Brisbane & Women's Hospital, Herston, Queensland, AustraliaSchool of Medicine, University of Queensland, Herston, Queensland, Australia
ALS patients have altered levels of metabolic proteins and adipokines relative to controls.
•
There is a positive correlation between the expression of plasma nerve growth factor relative to disease duration.
•
There is an inverse correlation between plasma glucagon and ALSFRS-R.
Abstract
Amyotrophic lateral sclerosis (ALS) is an adult-onset neurodegenerative disease characterized by the loss of upper cortical and lower motor neurons. ALS causes death within 2–5 years of diagnosis. Diet and body mass index influence the clinical course of disease, however there is limited information about the expression of metabolic proteins and fat-derived cytokines (adipokines) in ALS. In healthy controls and subjects with ALS, we have measured levels of proteins and adipokines that influence metabolism. We find altered levels of active ghrelin, gastric inhibitory peptide (GIP), pancreatic polypeptide (PP), lipocalin-2, plasminogen activator inhibitor-1 (PAI-1), interleukin-6 (IL-6) and 8 (IL-8), and tumor necrosis factor alpha (TNFα) in the plasma of ALS patients relative to controls. We also observe a positive correlation between the expression of plasma nerve growth factor (NGF) relative to disease duration, and an inverse correlation between plasma glucagon and the ALS functional rating scale-revised (ALSFRS-R). Further studies are required to determine whether altered expression of metabolic proteins and adipokines contribute to motor neuron vulnerability and how these factors act to modify the course of disease.
Amyotrophic lateral sclerosis (ALS) is a neurodegenerative disease that is characterized by the irreversible loss of upper cortical and lower motor neurons. ALS causes death within 2–5 years of diagnosis, generally due to respiratory failure [
] toxicity are reported to contribute to disease. Mutations in causative genes (e.g., SOD1 and TDP-43) are associated with familial ALS, but have also been observed in ALS cases with no known or obvious family history [
In the absence of curative therapies in ALS, it is important to identify factors that modify the progression of disease. Metabolic abnormalities have been described in ALS. These include hypermetabolism, which is observed in both familial and sporadic ALS patients [
]. The impact of altered metabolic balance on the course of disease is demonstrated by the finding that correcting for hypermetabolism improves survival in mouse models of ALS [
Whole body metabolism is regulated by a network of proteins and peptides. Given the evidence of metabolic abnormalities in ALS, this study aimed to determine whether ALS patients and age-matched healthy subjects differ in levels of metabolic proteins and adipokines that are involved in regulating metabolic homeostasis (targets assessed in this study are listed in Supplementary Table 1). This study also aimed to assess whether levels of these proteins and adipokines are associated with BMI, disease duration, and functional capacity (determined by the ALS functional rating scale-revised; ALSFRS-R) [
Ethical approval for the study was obtained from the Royal Brisbane and Women's Hospital (RBWH), The Wesley Hospital, and The University of Queensland ethics committees. The study was conducted in accordance with the principles set out in the Declaration of Helsinki. We enrolled 68 ALS patients (30.9% female, 69.1% male) from the RBWH Motor Neuron Disease Clinic between March 2010 and March 2014. For healthy controls, we enrolled 34 age-matched healthy volunteers from the RBWH, The Wesley Hospital, and The University of Queensland. Control subjects were not suffering from any disease condition and were not genetically related to the ALS patients recruited into this study. ALS patients had clinically definite or probable ALS (10.3% familial. 89.7% sporadic) according to the revised El Escorial criteria [
World Federation of Neurology Research Group on Motor Neuron D. El Escorial revisited: revised criteria for the diagnosis of amyotrophic lateral sclerosis.
Amyotroph. Lateral Scler. Other Motor Neuron Disord.2000; 1: 293-299
]. The clinical details of the patients are listed in Table 1. To be included in the study, patients were required to have onset of disease in the limbs. Patients with bulbar onset were excluded due to swallowing problems associated with bulbar symptoms. There was no significant difference between controls and ALS subjects in age or BMI. The functional status of ALS patients was evaluated using the ALSFRS-R [
Blood samples were collected between 1000 h and 1430 h into 4 ml BD Vacutainers® containing 1267.2 mg of the anti-coagulant K2EDTA (BD, Franklin Lakes, NJ, USA). Samples were centrifuged for 3 min at 3600 rpm. Plasma was collected, frozen on dry ice, and stored at −80 °C for further analysis. The duration of disease from onset of symptoms to the date of collection was recorded.
2.3 Multiplex assays
Plasma samples were treated with a protease inhibitor cocktail (Millipore, MA, USA; mM: 200 AEBSF, 0.16 bovine aprotinin, 10 bestatin, 3 E-64 protease inhibitor, 4 leupeptin, 2 pepstatin A), dipeptidyl peptidase 4 (DPP-IV; Millipore; 50 mM), and aprotinin (Sigma, MO, USA; 200 mg/ml) and phenylmethanesulfonylfluoride (PMSF; Sigma; 0.1 mg/ml). Magnetic multiplex assays were performed according to the manufacturer's instructions (Millipore). Targets on the human metabolic hormone assay (HMHMAG-34K) were: total amylin, c-peptide, GIP, ghrelin (active), glucagon, glucagon-like peptide-1 (GLP-1, active), PP and peptide YY (PYY). Targets on the human adipokine assays (HADK1MAG-61K and HADK2MAG-61K) were: adiponectin, hepatocyte growth factor (HGF), insulin, IL-6, IL-8, leptin, lipocalin-2, monocyte chemoattractant protein-1 (MCP-1), NGF, total PAI-1, resistin, and TNFα. The origin and biological function of these targets are summarized in Supplementary Table 1.
2.4 Statistical analysis
Data were analyzed using Prism 6.0c (Graphpad Software Inc., CA, USA). Normality of residuals (distribution) within control or ALS groups for all measures was assessed by Shapiro–Wilk W test. Differences between groups were determined by Mann–Whitney U test. The strength of associations between metabolic proteins and adipokines and BMI, ALSFRS-R scores and disease duration was determined by Spearman's rank correlation coefficient analysis. Data in tables are presented as mean ± SD. A P value of less than 0.05 denotes statistical significance.
3. Results
3.1 Expression of metabolic proteins and adipokines in ALS patients and controls
The expression of human metabolic proteins and adipokines in ALS patients and controls is presented in Table 2. For metabolic proteins, there was no difference in levels of total amylin, c-peptide, active GLP-1, glucagon, insulin, and PYY between controls and ALS patients. The plasma concentration of active ghrelin, GIP and PP was significantly lower in ALS patients than controls. For adipokines, there was no difference in levels of HGF, leptin, MCP-1, NGF, and resistin between controls and ALS patients. Circulating levels of adiponectin, IL-6, IL-8, lipocalin-2, PAI-1, and TNFα were significantly higher in ALS patients than in controls. Riluzole had no effect on the expression of any of the metabolic proteins and adipokines.
Table 2Comparison of the expression of metabolic markers and adipokines between control subjects and amyotrophic lateral sclerosis (ALS) patients.
Control
95% CI (range)
ALS
95% CI (range)
p
Metabolic factors
Amylin (pg/ml)
26.3 (14.4)
21.3–31.4
31.8 (36.3)
23.0–40.6
0.40
c-Peptide (ng/ml)
3.4 (1.9)
2.69–4.05
3.7 (5.5)
2.39–5.05
0.72
Ghrelin (pg/ml)
25.1 (28.2)
15.3–35.0
13.2 (11.2)
10.5–15.9
<0.01
GIP (ng/ml)
0.4 (0.3)
0.303–0.484
0.3 (0.3)
0.21–0.33
<0.01
GLP-1 (pg/ml)
30.3 (35.7)
345–627
28.5 (30.7)
230–379
0.80
Glucagon (pg/ml)
44.1 (34.1)
31.6–56.6
37.7 (24.1)
31.8–43.7
0.30
Insulin (pg/ml)
53.4 (46.4)
37.2–69.6
46.6 (51.7)
34.1–59.1
0.52
PP (pg/ml)
485.7 (404.0)
345–627
304.5 (305.5)
231–378
<0.01
PYY (pg/ml)
172.8 (123.2)
130–216
158.9 (107.3)
133–185
0.56
Adipokines
Adiponectin (mg/ml)
26.7 (22.7)
18.7–34.8
40.1 (45.4)
29.1–51.2
<0.05
HGF (pg/ml)
352.9 (211.3)
279–426
401.0 (260.2)
338–464
0.35
IL-6 (pg/ml)
3.0 (2.4)
2.07–3.81
9.0 (22.4)
3.59–145
<0.05
IL-8 (pg/ml)
3.1 (2.5)
2.23–3.97
4.6 (2.6)
3.93–5.21
<0.01
Leptin (ng/ml)
17.2 (27.2)
7.51–26.8
15.0 (27.5)
8.25–21.7
0.71
Lipocalin-2 (ng/ml)
104.3 (40.8)
90.1–119
123.9 (40.9)
114–134
<0.05
MCP-1 (pg/ml)
161.6 (76.1)
135–188
164.6 (76.8)
146–183
0.83
NGF (pg/ml)
8.0 (4.1)
6.59–9.45
10.3 (21.9)
4.98–15.6
0.55
PAI-1 (ng/ml)
55.5 (39.8)
41.6–69.3
95.9 (60.6)
81.2–111
<0.01
Resistin (ng/ml)
57.6 (37.0)
44.7–70.5
66.2 (39.3)
56.7–75.7
0.30
TNFα (pg/ml)
3.1 (1.4)
2.59–3.56
3.7 (1.4)
3.40–4.07
<0.01
denotes significance at p ≤ 0.05; denotes significance at p ≤ 0.01.
Of the factors that showed significant differences between ALS subjects and controls, ghrelin, adiponectin, IL-6, IL-8 and TNFα were differentially affected by gender within the ALS cohort (Supplementary Table 2). Ghrelin, which was reduced in the ALS group, was significantly lower in female ALS subjects. Adiponectin, which was higher in ALS subjects, was higher in female ALS subjects. IL6, IL8 and TNFα, which were higher in ALS subjects, were lower in female ALS subjects.
3.2 Correlation analysis of plasma metabolic proteins and adipokines with clinical features
3.2.1 Correlation of plasma metabolic proteins and adipokines with BMI
Of the 34 healthy subjects and 68 ALS patients, measures of BMI were available from 25 and 46 persons respectively. Supplementary Table 3 summarizes Spearman correlation analyses and comparison of fit for metabolic proteins and adipokines relative to BMI in controls and ALS patients. BMI was significantly correlated with levels of leptin, PAI-1, and MCP-1 in the controls (Fig. 1A–C ). For the ALS cohort, we observed significant positive correlation of plasma leptin, insulin and c-peptide with BMI (Fig. 1A, D and E). Plasma adiponectin showed a significant inverse correlation with BMI in ALS (Fig. 1F). We found no correlation between BMI and ALSFRS-R (Fig. 1G) or BMI and disease duration (Fig. 1H).
Fig. 1Correlation analysis of plasma metabolic proteins and adipokines in healthy controls and patients with amyotrophic lateral sclerosis (ALS). (A) BMI correlated with levels of leptin, (B) plasminogen activator inhibitor 1 (PAI-1), and (C) monocyte chemoattractant protein-1 (MCP-1) in the controls. For the ALS cohort, plasma levels of (A) leptin, (D) insulin, and (E) c-peptide increased relative to an increase in BMI. (F) Plasma adiponectin was inversely correlated with BMI in the ALS cohort. (G) Measures of BMI in ALS patients did not correlate with ALSFRS-R. (H) The inverse relationship between BMI and disease duration did not reach significance. (I) A significant inverse correlation exists between circulating levels of glucagon and ALSFRS-R, and (J) a positive correlation exists between circulating levels of NGF and disease duration.
3.2.2 Correlation of plasma metabolic proteins and adipokines with ALSFRS-R and disease duration
The correlation between levels of metabolic proteins and adipokines with ALSFRS-R scores and disease duration are shown in Supplementary Table 4. There was a significant inverse correlation between the expression of circulating levels of glucagon and ALSFRS-R in ALS patients (Fig. 1I). There was a significant positive correlation between circulating levels of NGF and disease duration in ALS patients (Fig. 1J).
4. Discussion
We present data showing differing levels of metabolic proteins and adipokines between ALS patients and healthy controls. The potential metabolic consequences of our findings specific to ALS are discussed below and are listed in Table 3.
Table 3Summary of the observed changes in metabolic proteins and adipokines in amyotrophic lateral sclerosis (ALS).
Hormone/peptide
Regulation in ALS
Relation to BMI in ALS
Relation to ALSFRS-R
Relation to disease duration
Possible metabolic consequences in ALS
Metabolic Factors
C-peptide
No change
Positive correlation
No correlation
No correlation
Impaired C-peptide release may underlie impaired insulin function and glucose homeostasis in ALS. Unlikely as secretion appears to be normal.
Ghrelin
Decreased
No correlation
No correlation
No correlation
Reduced hunger leading to reduced food intake and possibly weight loss.
GIP
Decreased
No correlation
No correlation
No correlation
Reduced insulin secretion resulting in reduced capacity to use glucose as energy; reduced lipogenesis resulting in an inability to accumulate fat mass.
Glucagon
No change
No correlation
Negative correlation
No correlation
Impaired capacity to increase circulating levels of glucose; impaired glucose tolerance resulting in reduced capacity to use glucose as energy. May not be reflected by circulating levels.
Insulin
No change
Positive correlation
No correlation
No correlation
Impaired insulin action thought to contribute to impaired glucose tolerance resulting in impaired capacity to use glucose as energy. May not be reflected by circulating levels.
PP
Decreased
No correlation
No correlation
No correlation
Reduced peripheral insulin action resulting in reduced capacity to use glucose as energy; decreased secretion of gastric juices resulting in gastrointestinal dysfunction.
Adipokines
Adiponectin
No change
Negative correlation
No correlation
No correlation
Increased risk of metabolic derangements associated with increased fat mass (eg. insulin resistance).
IL-6
Increased
No correlation
No correlation
No correlation
May contribute to the development of insulin resistance resulting in reduced capacity to use glucose as energy.
IL-8
Increased
No correlation
No correlation
No correlation
May contribute to the development of insulin resistance resulting in reduced capacity to use glucose as energy.
Leptin
No change
Positive correlation
No correlation
No correlation
Increased satiety relative to fat mass resulting in reduced food intake as a means to modulate energy balance.
Lipocalin-2
Increased
No correlation
No correlation
No correlation
Increased total energy metabolism and increased breakdown of fat leading to depletion of fat mass and subsequent weight loss and reduced BMI.
NGF
No change
No correlation
No correlation
Positive correlation
Increase relative to disease duration may reflect reduced food intake.
PAI-1
Increased
No correlation
No correlation
No correlation
May contribute to the development of insulin resistance resulting in reduced capacity to use glucose as energy.
TNFα
Increased
No correlation
No correlation
No correlation
Induction of lipolysis resulting in the breakdown and mobilization of fat stores and subsequent weight loss and reduced BMI.
]. Consistent with reduced food intake in ALS, we observe lower levels of GIP and PP in ALS patients. GIP stimulates insulin secretion in response to food intake [
Glucagon-like peptide-1 (7–36)amide and glucose-dependent insulinotropic polypeptide secretion in response to nutrient ingestion in man: acute post-prandial and 24-h secretion patterns.
]. Reduced levels of GIP and PP in our ALS cohort may provide a mechanism for the diminished peripheral insulin action and glucose intolerance observed in ALS [
We observed increased expression of a number of adipokines that are linked to metabolic disease. Adiponectin regulates glucose and fatty acid oxidation [
Association between measures of insulin sensitivity and circulating levels of interleukin-8, interleukin-6 and tumor necrosis factor-alpha. Effect of weight loss in obese men.
Visceral fat accumulation is an important determinant of PAI-1 levels in young, nonobese men and women: modulation by cross-sex hormone administration.
]. Therefore, the increased expression of adiponectin, IL-6, IL-8, lipocalin-2, PAI-1 and TNFα that we observe in ALS may contribute to the reported occurrence of glucose intolerance and insulin resistance [
] that is seen in ALS. Nonetheless, the mechanism for the increase in the expression of these adipokines might also be inherent to pathologies associated with ALS. Indeed, it is possible that IL-6, IL-8, and TNFα originate from the activation of innate immunity [
] are positively associated with BMI, the correlation between levels of PAI-1 and MCP-1 with BMI in our control cohort is to be expected. In contrast, we observed no association between PAI-1 and MCP-1 with BMI in our ALS cohort. PAI-1 and MCP-1 are associated with changes in the distribution of or mass of visceral fat [
Visceral fat accumulation is an important determinant of PAI-1 levels in young, nonobese men and women: modulation by cross-sex hormone administration.
] might underlie the greater variability in circulating levels of PAI-1 and MCP-1 and consequently the lack of association between PAI-1 and MCP-1 and BMI in the ALS cohort.
Insulin and c-peptide are associated with increased adiposity and BMI. Thus, the relationship between insulin and BMI in subgroup of control and ALS subjects that were BMI-matched, and the lack of deviation in this relationship between the two cohorts is to be expected. As c-peptide is co-secreted with insulin, our observations that BMI correlated with c-peptide levels in the ALS cohort mirror the positive correlation between insulin and BMI.
While adiponectin is inversely associated with fat mass [
Plasma adiponectin concentration is associated with skeletal muscle insulin receptor tyrosine phosphorylation, and low plasma concentration precedes a decrease in whole-body insulin sensitivity in humans.
], the lack of relationship between circulating levels of adiponectin and BMI are in line with observations that BMI does not independently affect plasma adiponectin levels in healthy lean and obese individuals [
]. The majority of our control and ALS patients fell within the healthy to overweight BMI range, yet we observed a negative association between adiponectin and BMI in ALS patients. Since adiponectin levels are correlated with the development of metabolic complications of adiposity [
We found that lower levels of circulating glucagon were associated with higher ALSFRS-R scores. This suggests that an increase in disease severity in ALS patients is correlated with increased circulating glucagon levels. Indeed, given that impaired glucose homeostasis in ALS could occur as a consequence of higher levels of circulating glucagon [
In our ALS patients, NGF levels correlated with disease duration. This may be related to increased fat mass, as is seen in obesity and metabolic syndrome [
]. As we observed no relationship between BMI and disease duration in our ALS cohort, it seems unlikely that the association of NGF with disease duration is linked with metabolic derangements associated with fat mass. Rather, increased NGF in patients with longer disease duration may underlie reduced food intake, since NGF suppresses food intake [
Neurotrophic and metabotrophic potential of nerve growth factor and brain-derived neurotrophic factor: linking cardiometabolic and neuropsychiatric diseases.
While the ALS subjects in this study were typical of a clinical population in that approximately 10% had familial ALS, we excluded patients with bulbar onset disease to remove any metabolic disturbance that might occur from malnutrition due to an inability to eat. Although the patients were studied at the same time of day, we did not control for food intake. Meal patterning influences the majority of metabolic proteins. Although we observed clear and consistent differences between ALS patients and healthy controls, our data need to be confirmed in a study that strictly controls for food intake, and that is extended to patients with bulbar onset disease.
Under normal physiological conditions, the expression of metabolic proteins and adipokines varies relative to BMI and age. Given that BMI values were only matched for a subset of our control and ALS populations, we cannot exclude that the altered expression of metabolic proteins and adipokines observed in this study might be related to BMI. However, in contrast to previous reports [
], we found no correlation between BMI and disease severity or disease duration in our ALS patients with known BMI. Given that our measures do not account for the progressive weight loss that is seen in ALS patients during disease progression, long-term serial assessment of the relationship between BMI and ALSFRS-R or disease duration is also needed to assess associations between BMI and disease severity, and BMI and disease progression.
Gender impacts the prevalence and severity of ALS [
], and thus gender-effects may contribute to the differential expression of metabolic proteins and adipokines between ALS patients and controls. Levels of amylin, ghrelin, adiponectin, HGF, IL-6, IL-8, resistin and TNFα differed between male and female ALS patients. Of these, amylin, ghrelin, HGF, and IL-8 levels varied from anticipated gender specific differences that would normally be observed [
IL-8 plasma concentrations and the risk of future coronary artery disease in apparently healthy men and women: the EPIC-Norfolk prospective population study.
], it is possible that the heterogeneity of ALS may lead to varying degrees of metabolic dysfunction between patients with different ALS genotypes and phenotypes. Thus, there is a need for studying metabolic disturbances across the whole spectrum of ALS.
In summary, we observe a number of alterations in the expression of metabolic proteins and adipokines between ALS patients and BMI matched controls. These data, in combination with current clinical observations of the impact of altered metabolic balance in ALS, provide evidence to substantiate the existence of metabolic responses in ALS, and shed light on altered homeostatic processes in this complex disease. Future studies are needed to determine whether altered expression of metabolic proteins and adipokines contribute to motor neuron vulnerability, or whether/how these factors act to modify the course of disease.
Acknowledgments
We dedicate this manuscript to the memory of Mr Scott Sullivan and Mr Trevor Overfield. This work was funded by a Grant-in-Aid to PAM, STN and FJS from the Motor Neurone Disease Research Institute of Australia (MNDRIA), Royal Brisbane & Women's Hospital Foundation grants to RDH, and The MND and Me Foundation funds to RDH. STN acknowledges the support of an MNDRIA Bill Gole Fellowship and The University of Queensland (2012-2015), and the support of a Scott Sullivan MND Research Fellowship (2015-2018) funded by the Queensland Brain Institute, The Royal Brisbane & Women’'s Hospital Foundation, and The MND and Me Foundation. STN acknowledges funding support from The School of Biomedical Sciences, University of Queensland. We extend our gratitude to Ms Nicole Hutchinson, Ms Helen Woodhouse and Ms Susan Heggie for their assistance with collecting patient blood and patient clinical history. Finally, we extend our sincerest gratitude to all ALS patients and healthy volunteers who participated in this study.
Supp Table 1. Summary of source and biological role of metabolic factors and adipose specific or enriched proteins (adipokines) assessed.
Supp Table 2, Comparison of the expression of metabolic markers and adipokines between male and female amyotrophic lateral sclerosis (ALS) patients when matched for body mass index (BMI).
Supp Table 3. Spearman correlation analysis relative to body mass index (BMI) and comparison of fit for metabolic factors and adipokines in control subjects and amyotrophic lateral sclerosis (ALS) patients.
Supp Table 4. Spearman correlation analysis of metabolic factors and adipokines relative to amyotrophic lateral sclerosis functional rating scale-revised (ALSFRS-R) scores and disease duration in patients.
References
Rothstein J.D.
Current hypotheses for the underlying biology of amyotrophic lateral sclerosis.
World Federation of Neurology Research Group on Motor Neuron D. El Escorial revisited: revised criteria for the diagnosis of amyotrophic lateral sclerosis.
Amyotroph. Lateral Scler. Other Motor Neuron Disord.2000; 1: 293-299
Glucagon-like peptide-1 (7–36)amide and glucose-dependent insulinotropic polypeptide secretion in response to nutrient ingestion in man: acute post-prandial and 24-h secretion patterns.
Association between measures of insulin sensitivity and circulating levels of interleukin-8, interleukin-6 and tumor necrosis factor-alpha. Effect of weight loss in obese men.
Visceral fat accumulation is an important determinant of PAI-1 levels in young, nonobese men and women: modulation by cross-sex hormone administration.
Plasma adiponectin concentration is associated with skeletal muscle insulin receptor tyrosine phosphorylation, and low plasma concentration precedes a decrease in whole-body insulin sensitivity in humans.
Neurotrophic and metabotrophic potential of nerve growth factor and brain-derived neurotrophic factor: linking cardiometabolic and neuropsychiatric diseases.
IL-8 plasma concentrations and the risk of future coronary artery disease in apparently healthy men and women: the EPIC-Norfolk prospective population study.
denotes significance at p ≤ 0.05; 
00394-9&id=gr1.jpg){kind=link}