- Email: [email protected]
Phospholipid transfer protein activity and incident type 2 diabetes mellitus
Clinica Chimica Acta 439 (2015) 38–41
Contents lists available at ScienceDirect
Clinica Chimica Acta journal homepage: www.elsevier.com/locate/clinchim
Phospholipid transfer protein activity and incident type 2 diabetes mellitus☆ Ali Abbasi a,b,c, Geesje M. Dallinga-Thie d, Robin P.F. Dullaart c,⁎ a MRC Epidemiology Unit, University of Cambridge School of Clinical Medicine, Institute of Metabolic Science, Cambridge Biomedical Campus, Addenbrooke's Hospital, P.O. Box 285, Cambridge CB2 0QQ, UK b Department of Epidemiology, University of Groningen, University Medical Center Groningen, Groningen, the Netherlands c Department of Internal Medicine, University of Groningen, University Medical Center Groningen, Groningen, the Netherlands d Department of Experimental Vascular Medicine, Academic Medical Center, Amsterdam, the Netherlands
a r t i c l e
i n f o
Article history: Received 15 September 2014 Received in revised form 29 September 2014 Accepted 30 September 2014 Available online 7 October 2014 Keywords: Epidemiology Phospholipid transfer protein activity Risk factor Type 2 diabetes mellitus
a b s t r a c t Background: The plasma activity of phospholipid transfer protein (PLTP), which has multifaceted functions in lipoprotein metabolism and in inflammatory responses, is elevated in insulin resistant conditions. We determined the association of plasma PLTP activity with incident type 2 diabetes mellitus (T2DM). Methods: Plasma PLTP activity was determined using a liposome vesicle–HDL system in 218 men without T2DM at baseline. We used logistic regression models to establish odds ratios (ORs) for incident T2DM. Results: Twenty four men developed T2DM over 9.4-year follow-up. Plasma PLTP activity was higher in incident T2DM cases (p = 0.009). We observed 82% higher odds for T2DM per 1-SD increase in PLTP activity. Multivariable modeling showed that the association of PLTP activity with T2DM was independent of clinical risk factors including age, and either the metabolic syndrome, individual metabolic syndrome components, total cholesterol, HOMA-IR or albuminuria (ORs ranging from 1.64 (95% CI 1.03–2.66) to 1.87 (1.19–3.010)). Conclusions: Elevated plasma PLTP activity may predict an increased risk of T2DM in men. © 2014 Elsevier B.V. All rights reserved.
1. Introduction Human plasma contains phospholipid transfer protein (PLTP) which has multifaceted functions in lipoprotein metabolism, including transfer of phospholipids to high density lipoproteins (HDL), conversion of mature α-HDL towards larger and smaller HDL particles, and stimulation of very low density lipoprotein (VLDL) secretion by the liver [1]. PLTP is also involved in inflammatory responses and in antioxidant defense. Evidence has accumulated in support of the hypothesis that high plasma PLTP activity may predict (subclinical) atherosclerotic cardiovascular disease (CVD) in rodents and in humans [1–4]. Several compounds have been recently developed which inhibit PLTP activity and decrease hepatic apolipoprotein B secretion in vitro [2]. Interestingly, plasma glucose is decreased in PLTP deficient mouse models [5]. Moreover, it has consistently been observed that plasma PLTP activity is increased in insulin resistant states, such as type 2 diabetes mellitus (T2DM) and obesity [2]. Although these findings raise the hypothesis that PLTP may be implicated in the development of T2DM, the association of plasma PLTP activity with incident T2DM has yet ☆ Disclosure Statement: The authors have nothing to disclose. ⁎ Corresponding author at: Department of Internal Medicine, University Medical Center Groningen, P.O. Box 30.001, Groningen 9700 RB, the Netherlands. E-mail address: [email protected] (R.P.F. Dullaart).
http://dx.doi.org/10.1016/j.cca.2014.09.035 0009-8981/© 2014 Elsevier B.V. All rights reserved.
not been tested. The aim of the present study was to determine the extent to which plasma PLTP activity predicts incident T2DM in humans.
2. Materials and methods 2.1. Study design and participants The study was approved by the Medical Ethics Committee of the University Medical Center Groningen, Groningen, the Netherlands, and all participants gave written informed consent. We prospectively analyzed data on 218 men without diabetes at baseline, who were followed from 1996–97 until January 1st 2010, using data from a previously reported nested case–control study on CVD among men participating in the Prevention of REnal and Vascular ENd stage Disease (PREVEND) study (n = 8,592), Groningen, the Netherlands [6]. In brief, original selection was men without diabetes and CVD at baseline who developed CVD before initial census date (31 December 2003 or 31 December 2002 until which date information regarding specific causes of death was available) and a group of randomly selected men who did not experience CVD during this follow-up period [6]. In 218 of the initially selected 227 men, baseline samples for PLTP activity assay were available and new-onset T2DM could be ascertained during follow-up.
A. Abbasi et al. / Clinica Chimica Acta 439 (2015) 38–41
2.2. Primary outcome Incident T2DM cases were ascertained if one or more of the following conditions were fulfilled during follow-up: a physician diagnosed T2DM as indicated by self-report, fasting plasma glucose ≥7.0 mmol/l, a random sample plasma glucose ≥ 11.1 mmol/l or initiation of glucose-lowering medication as retrieved from a central pharmacy registry [7]. 2.3. Plasma PLTP activity and other laboratory measurements PLTP activity (in arbitrary units [AU] relative to pool plasma) was measured with a liposome vesicle–HDL system in EDTA-anticoagulated plasma samples that were stored at − 80 °C until assay (intra-assay coefficient of variation b 5%) [3]. All other laboratory assays were also carried out in the baseline samples. Glucose, lipids, lipoproteins and insulin were assayed as described [7]. 2.4. Statistical analysis Statistical analyses were performed using IBM SPSS or R-2.13.1 for Windows (http://cran.r-project.org/). Data are given in mean SD, median (interquartile ranges) or numbers. Between group differences were compared by unpaired t tests, Mann–Whitney U tests and chi-square analysis where appropriate. Univariate relationships were determined by Spearman's rank correlation coefficients (r). Logistic regression was used to test the association of PLTP activity with incident T2DM, since new-onset diabetes has been ascertained periodically in the PREVEND cohort, i.e. at the regular screening visits. Odds ratios (ORs) are given with 95% confidence intervals (CIs). Given the rather low number of subjects with incident T2DM, multivariable adjustment was done for maximally 2 variables together. The additive value of PLTP activity for T2DM prediction was estimated in terms of change in the C-statistic and reclassification (integrated discrimination improvement (IDI)) using the presence of the metabolic syndrome or the Data from the Epidemiological Study on the Insulin Resistance Syndrome (DESIR) risk score (a sex-specific diabetes prediction model, validated in PREVEND) as reference [8]. Two-sided pvalues b 0.05 were considered statistically significant. 3. Results During a follow-up of 9.4 ± 0.7 (mean ± SD) years, 24 participants (11.0%) developed T2DM. Plasma PLTP activity was normally distributed (Kolmogorov–Smirnov test: p = 0.20), and was higher in the individuals who developed T2DM compared to non-convertors (106.4 ± 8.7 vs. 100.3 ± 10.8 AU, p = 0.009; Table 1). In all participants combined, PLTP activity was correlated positively with waist circumference (r = 0.145, p = 0.03), the DESIR risk score (r = 0.160, p = 0.018), plasma insulin (r = 0.216, p = 0.001), insulin resistance (homeostasis model assessment (HOMA-IR)) (r = 0.201, p = 0.003), total cholesterol (r = 0.138, p = 0.042) and triglycerides (r = 0.211, p = 0.002), but not with glucose (r = 0.036, p = 0.599). We observed crude 82% higher odds for T2DM per 1-SD increase in PLTP activity (Table 1). The association of PLTP activity with T2DM remained significant after adjustment for age and either the presence of the metabolic syndrome, individual metabolic syndrome components, total cholesterol, the DESIR risk score, HOMA-IR, the case–control design or albuminuria. Corresponding odds ratios (95% CI) for incident T2DM ranged from 1.64 (1.03–2.66) to 1.87 (1.19–3.01) per 1-SD increase in PLTP activity (Table 2). Addition of PLTP activity to reference models of either the metabolic syndrome or the DESIR prediction model resulted in improvement in the C-statistic or reclassification (IDI). The greatest C-statistic and IDI were achieved when we added PLTP activity along with glucose to the DESIR model (Table 3).
39
Table 1 Baseline clinical and laboratory characteristics and the relationship of plasma phospholipid transfer protein activity with incident type 2 diabetes mellitus. Variablesa
Non-convertor (n = 194)
Incident type 2 diabetes mellitus (n = 24)
Age (years) Incident CVD (number; %) Smoking (number; %) BMI (kg/m2) Waist circumference (cm) Systolic blood pressure (mm Hg) Diastolic blood pressure (mm Hg) Hypertension (number; %) Metabolic syndrome (number; %) DESIR risk score UAE (mg/24 h) Glucose (mmol/l) Insulin (pmol/l)
51.8 ± 12.4 98 (50.5) 90 (46.4) 26.4 ± 3.4 94 ± 10 137 ± 20
57.6 ± 10.1 14 (58.3) 10 (41.7) 28.2 ± 4.5 100 ± 10 144 ± 19
0.03 0.48 0.66 0.02 0.007 0.14
79 ± 9
81 ± 6
0.32
89 (45.6) 46 (23.7)
15 (62.9) 14 (58.3)
0.12 b0.001
2.82 ± 1.21 12.1 (7.5–22.7) 4.84 ± 0.58 47.4 (35.0–73.2) 1.7 (1.2–2.6) 5.88 ± 1.14 1.14 ± 0.36 1.31 (0.90–1.92) 100.3 ± 10.8
3.33 ± 1.09 20.2 (10.2–35.9) 5.47 ± 0.69 73.2 (50.4–115.2) 2.9 (2.0–4.8) 6.24 ± 0.88 1.05 ± 0.34 1.81 (1.45–2.67) 106.4 ± 8.7
0.02 0.03 b0.001 0.001
HOMA-IR Total cholesterol (mmol/l) HDL-cholesterol (mmol/l) Triglycerides (mmol/l) PLTP activity (AU)
p-Value
b0.001 0.13 0.23 0.003 0.009
The metabolic syndrome was defined according to the revised National Cholesterol Education Program's Adult Treatment Panel III report with at least three of the following criteria: 1) waist circumference N102 cm, 2) blood pressure ≥130/85 mm Hg or treatment for hypertension, 3) fasting triglycerides ≥1.7 mmol/l, 4) high-density lipoprotein (HDL) cholesterol ≤1.0 mmol/l, and 5) fasting plasma glucose ≥5.6 mmol/l or treatment for type 2 diabetes mellitus. The DESIR risk score was calculated as a sum of waist circumference, current smoking and hypertension as follows [8]: waist circumference b80 cm: 0; 80–89 cm: 1; 90–99 cm: 2; ≥100 cm: 3; current smoking: 1; hypertension: 1 (total score: 0–5). The absolute risk of type 2 diabetes is calculated as: exp(α + β × linear predictor)/(1 + exp(α + β × linear predictor)), where, α is the calibration intercept, β is the calibration slope, and linear predictor (if men) = −10.45 + 0.72 (if current smoker, otherwise 0) + 0.081 × waist circumference (cm) + 0.50 (if hypertension, otherwise 0). HOMA-IR was calculated as fasting plasma insulin [in mU/l] × fasting plasma glucose [in mmol/l] divided by 22.5. a Data are shown as mean ± SD or as median (interquartile ranges) for continuous variables which were compared by independent t tests or Mann–Whitney U tests where appropriate. Categorical variables are shown as frequencies (percentage), and were compared by χ2 test. Abbreviations: AU, arbitrary units; CVD, cardiovascular disease; BMI, body mass index (the weight in kilogram divided by the square of the height in meters); DESIR, Data from the Epidemiological Study on the Insulin Resistance Syndrome; HDL, high-density lipoproteins; LDL, low-density lipoproteins; HOMA-IR, homeostasis model assessment; PLTP, phospholipid transfer protein; UAE, urinary albumin excretion.
4. Discussion Our prospective study supports the hypothesis that higher plasma PLTP activity is associated with an increased risk for developing T2DM in men. Remarkably, the strength of association of PLTP activity with incident T2DM remained essentially similar when taking account of established diabetes risk factors, HDL cholesterol, triglycerides, total cholesterol, HOMA-IR or albuminuria. Of potential interest, diabetes risk prediction tended to improve when PLTP activity was added to reference models of the metabolic syndrome or the DESIR risk score together with glucose. Our findings suggest that the putative relationship between plasma PLTP activity and glucose homeostasis is at least in part independent from conventional diabetes risk factors and plasma lipid levels. PLTP could influence glucose homeostasis via several yet to be more precisely delineated processes. PLTP deficiency has been recently shown to enhance pancreatic glucose-stimulated insulin secretion and to increase peripheral glucose disposal after lipopolysaccharide administration-induced endotoxemia in mice [9]. Additionally, PLTP gene expression is regulated by liver X receptor-, farnesoid X-activated receptor- and peroxisome proliferator activated receptor-dependent mechanisms, which are also involved in glucose homeostasis [10].
40
A. Abbasi et al. / Clinica Chimica Acta 439 (2015) 38–41
Table 2 Predictive value of plasma phospholipid transfer protein (PLTP) activity for incident type 2 diabetes mellitus by logistic regression analysis. Logistic regression models
OR (95% CI) per 1-SD PLTP activity
p-value
Crude Model 1 Model 2 Model 3 Model 4 Model 5 Model 6 Model 7 Model 8 Model 9 Model 10 Model 11 Model 12
1.82 (1.17–2.91) 1.84 (1.18–2.95) 1.79 (1.13–2.91) 1.87 (1.19–3.01) 1.82 (1.17–2.93) 1.82 (1.15–2.94) 1.86 (1.20–2.99) 1.68 (1.07–2.70) 1.81 (1.16–2.91) 1.74 (1.11–2.80) 1.64 (1.03–2.66) 1.84 (1.18–2.95) 1.84 (1.18–2.96)
0.010 0.008 0.016 0.008 0.010 0.012 0.007 0.028 0.010 0.018 0.039 0.009 0.009
Sources of funding This work was supported by the Netherlands Heart Foundation (2001.005), Dutch Diabetes Research Foundation, Dutch Kidney Foundation, the Netherlands Organization for Scientific Research project (NWO) and the Medical Research Council UK (grant reference no. MC_U106179471). AA is supported by a Rubicon grant from the Netherlands Organization for Scientific Research (NWO). Disclosures
Odds ratios (ORs) with 95% confidence intervals (CI) are expressed per SD increment in PLTP activity Model 1: adjusted for age; model 2: adjusted for age plus the presence of the metabolic syndrome; model 3: adjusted for age plus hypertension criterion; model 4: adjusted for age plus waist circumference criterion; model 5: adjusted for age plus glucose criterion; model 6: adjusted for age plus high density lipoprotein cholesterol criterion; model 7: adjusted for age plus triglyceride criterion; model 8: adjusted for age plus total cholesterol; model 9: adjusted for age plus “Data from the Epidemiological Study on the Insulin Resistance Syndrome” (DESIR) score; model 10: adjusted for age plus homeostasis model assessment; model 11: adjusted for age plus case–control study design; model 12: adjusted for age plus urinary albumin excretion.
PLTP does not only affect key pathways involved in lipoprotein metabolism, including HDL remodeling resulting in enhanced hepatic VLDL secretion and in smaller-sized HDL particles, but may also influence the quality of plasma lipoproteins [1,2]. Furthermore, PLTP exerts several actions at the cellular level, including effects on the nuclear factor-κB pathway in macrophages with functional consequences for vascular biology and inflammation [1,2]. Of additional interest, common genetic variations in PLTP associate with glycemic traits [11]. Remarkably, in contrast to the positive relationship between higher triglycerides, lower HDL cholesterol and higher glucose levels, genetic predisposition to dyslipidemia may associate with a pleiotropic lowering influence on glycemic traits [11]. Limitations of the present study include its small sample size and the lack of power to explore potential interactions. Furthermore, the baseline difference in plasma PLTP activity between men who developed T2DM compared to men who did not was modest (6.1%), possibly limiting its clinical utility. Notably, the cumulative incidence of T2DM in this cohort was quite comparable to that of men without diabetes at baseline in the PREVEND cohort [12]. The relationship of PLTP with diabetes development needs to be replicated, particularly in women and in non-white populations. However, no large cohorts are available with plasma PLTP activity measurements using the original liposome assay. Inherent to an observational study, we could not address the nature of relationship and reverse causation. In conclusion, our results underscore the hypothesis that plasma PLTP activity may serve as a biomarker of incident T2DM, and may provide a rationale to determine possible effects of pharmacological PLTP inhibition on glucose homeostasis.
None of the study sponsors had a role in the study design, data collection, analysis and interpretation, report writing, or the decision to submit the report for publication. All authors declare no conflict of interest, no financial relationships with any organizations that might have an interest in the submitted work in the previous three years, no other relationships or activities that could appear to have influenced the submitted work. All authors declare that there is no duality of interest associated with their contribution to this manuscript. Contribution statement AA and RPD conceived and designed the study. AA performed statistical analysis for the study. AA, RPD and GD-T contributed to materials and analysis tools or to acquisition of data, data analysis and interpretation. AA and RPD wrote the first draft of the manuscript. AA, RPD and GD-T contributed to interpretation of data, critical revision of the manuscript for important intellectual content. AA, RPD and GD-T contributed to final approval of the version to be published. AA and RPD (who supervised the study) had full access to all of the data in the study and were guarantors of the study. Acknowledgments Prof. J.L. Hillege, Department of Cardiology, University of Groningen and University Medical Center Groningen, provided advice in the design of the nested case–control study. The laboratory work of J.J. Duker and J. van der Wal–Haneveld is appreciated. Prof. L.T.W. de Jong-van den Berg and dr. S.T. Visser Department of Social Pharmacy, Pharmacoepidemiology and Pharmacotherapy, Groningen University Institute for Drug Exploration, University of Groningen, University Medical Center Groningen provided the data on pharmacyregistered use of anti-diabetic medications. References [1] Tzotzas T, Desrumaux C, Lagrost L. Plasma phospholipid transfer protein (PLTP): review of an emerging cardiometabolic risk factor. Obes Rev 2009;10:403–11. [2] Dullaart RPF, van Tol A, Dallinga-Thie GM. Phospholipid transfer protein, an emerging cardiometabolic risk marker: is it time to intervene? Atherosclerosis 2013;228: 38–41. [3] Vergeer M, Boekholdt SM, Sandhu MS, et al. Genetic variation at the phospholipid transfer protein locus affects its activity and high-density lipoprotein size and is a novel marker of cardiovascular disease susceptibility. Circulation 2010;122:470–7.
Table 3 Contribution of PLTP activity to change in C-statistic and integration discrimination improvement (IDI).
Metabolic syndrome (yes/no)a Metabolic syndrome plus PLTP activity DESIR scoreb DESIR plus PLTP activity DESIR plus glucose DESIR plus glucose plus PLTP activity a b
C-statistic for models
p-Value for change in C-statistic
IDI for extended models
0.678 (0.575–0.785) 0.756 (0.663–0.848) 0.641 (0.528–0.754) 0.697 (0.594–0.801) 0.771 (0.641–0.879) 0.802 (0.711–0.894)
Reference 0.008 Reference 0.16 0.04 0.006
Reference 0.028 Reference 0.026 0.101 0.128
The presence of the metabolic syndrome was used as reference category. The “Data from the Epidemiological Study on the Insulin Resistance Syndrome” (DESIR) risk score was used as reference category.
p-Value for IDI 0.08 0.03 0.001 b0.001
A. Abbasi et al. / Clinica Chimica Acta 439 (2015) 38–41 [4] Moerland M, Samyn H, van Gent T, et al. Acute elevation of plasma PLTP activity strongly increases pre-existing atherosclerosis. Arterioscler Thromb Vasc Biol 2008;28:1277–82. [5] Schlitt A, Liu J, Yan D, Mondragon-Escorpizo M, Norin AJ, Jiang XC. Antiinflammatory effects of phospholipid transfer protein (PLTP) deficiency in mice. Biochim Biophys Acta 2005;1733:187–91. [6] Borggreve SE, Hillege HL, Dallinga-Thie GM, et al. High plasma cholesteryl ester transfer protein levels may favour reduced incidence of cardiovascular events in men with low triglycerides. Eur Heart J 2007;28:1012–8. [7] Abbasi A, Corpeleijn E, Gansevoort RT, et al. Role of HDL cholesterol and estimates of HDL particle composition in future development of type 2 diabetes in the general population: the PREVEND study. J Clin Endocrinol Metab 2013;98:E1352–9. [8] Balkau B, Lange C, Fezeu L, et al. Predicting diabetes: clinical, biological, and genetic approaches: data from the Epidemiological Study on the Insulin Resistance Syndrome (DESIR). Diabetes Care 2008;31:2056–61.
41
[9] Nguyen AT, Mandard S, Dray C, et al. Lipopolysaccharides-mediated increase in glucose-stimulated insulin secretion: involvement of the GLP-1 pathway. Diabetes 2014;63:471–82. [10] Lakomy D, Rebe C, Sberna AL, et al. Liver X receptor-mediated induction of cholesteryl ester transfer protein expression is selectively impaired in inflammatory macrophages. Arterioscler Thromb Vasc Biol 2009;29:1923–9. [11] Li N, van der Sijde MR, Study LC, et al. Pleiotropic effects of lipid genes on plasma glucose, HbA1c and HOMA-IR levels. Diabetes 2014;63:3149–58. [12] Abbasi A, Corpeleijn E, Meijer E, et al. Sex differences in the association between plasma copeptin and incident type 2 diabetes: the Prevention of Renal and Vascular Endstage Disease (PREVEND) study. Diabetologia 2012;55 [1963-197].
Contents lists available at ScienceDirect
Clinica Chimica Acta journal homepage: www.elsevier.com/locate/clinchim
Phospholipid transfer protein activity and incident type 2 diabetes mellitus☆ Ali Abbasi a,b,c, Geesje M. Dallinga-Thie d, Robin P.F. Dullaart c,⁎ a MRC Epidemiology Unit, University of Cambridge School of Clinical Medicine, Institute of Metabolic Science, Cambridge Biomedical Campus, Addenbrooke's Hospital, P.O. Box 285, Cambridge CB2 0QQ, UK b Department of Epidemiology, University of Groningen, University Medical Center Groningen, Groningen, the Netherlands c Department of Internal Medicine, University of Groningen, University Medical Center Groningen, Groningen, the Netherlands d Department of Experimental Vascular Medicine, Academic Medical Center, Amsterdam, the Netherlands
a r t i c l e
i n f o
Article history: Received 15 September 2014 Received in revised form 29 September 2014 Accepted 30 September 2014 Available online 7 October 2014 Keywords: Epidemiology Phospholipid transfer protein activity Risk factor Type 2 diabetes mellitus
a b s t r a c t Background: The plasma activity of phospholipid transfer protein (PLTP), which has multifaceted functions in lipoprotein metabolism and in inflammatory responses, is elevated in insulin resistant conditions. We determined the association of plasma PLTP activity with incident type 2 diabetes mellitus (T2DM). Methods: Plasma PLTP activity was determined using a liposome vesicle–HDL system in 218 men without T2DM at baseline. We used logistic regression models to establish odds ratios (ORs) for incident T2DM. Results: Twenty four men developed T2DM over 9.4-year follow-up. Plasma PLTP activity was higher in incident T2DM cases (p = 0.009). We observed 82% higher odds for T2DM per 1-SD increase in PLTP activity. Multivariable modeling showed that the association of PLTP activity with T2DM was independent of clinical risk factors including age, and either the metabolic syndrome, individual metabolic syndrome components, total cholesterol, HOMA-IR or albuminuria (ORs ranging from 1.64 (95% CI 1.03–2.66) to 1.87 (1.19–3.010)). Conclusions: Elevated plasma PLTP activity may predict an increased risk of T2DM in men. © 2014 Elsevier B.V. All rights reserved.
1. Introduction Human plasma contains phospholipid transfer protein (PLTP) which has multifaceted functions in lipoprotein metabolism, including transfer of phospholipids to high density lipoproteins (HDL), conversion of mature α-HDL towards larger and smaller HDL particles, and stimulation of very low density lipoprotein (VLDL) secretion by the liver [1]. PLTP is also involved in inflammatory responses and in antioxidant defense. Evidence has accumulated in support of the hypothesis that high plasma PLTP activity may predict (subclinical) atherosclerotic cardiovascular disease (CVD) in rodents and in humans [1–4]. Several compounds have been recently developed which inhibit PLTP activity and decrease hepatic apolipoprotein B secretion in vitro [2]. Interestingly, plasma glucose is decreased in PLTP deficient mouse models [5]. Moreover, it has consistently been observed that plasma PLTP activity is increased in insulin resistant states, such as type 2 diabetes mellitus (T2DM) and obesity [2]. Although these findings raise the hypothesis that PLTP may be implicated in the development of T2DM, the association of plasma PLTP activity with incident T2DM has yet ☆ Disclosure Statement: The authors have nothing to disclose. ⁎ Corresponding author at: Department of Internal Medicine, University Medical Center Groningen, P.O. Box 30.001, Groningen 9700 RB, the Netherlands. E-mail address: [email protected] (R.P.F. Dullaart).
http://dx.doi.org/10.1016/j.cca.2014.09.035 0009-8981/© 2014 Elsevier B.V. All rights reserved.
not been tested. The aim of the present study was to determine the extent to which plasma PLTP activity predicts incident T2DM in humans.
2. Materials and methods 2.1. Study design and participants The study was approved by the Medical Ethics Committee of the University Medical Center Groningen, Groningen, the Netherlands, and all participants gave written informed consent. We prospectively analyzed data on 218 men without diabetes at baseline, who were followed from 1996–97 until January 1st 2010, using data from a previously reported nested case–control study on CVD among men participating in the Prevention of REnal and Vascular ENd stage Disease (PREVEND) study (n = 8,592), Groningen, the Netherlands [6]. In brief, original selection was men without diabetes and CVD at baseline who developed CVD before initial census date (31 December 2003 or 31 December 2002 until which date information regarding specific causes of death was available) and a group of randomly selected men who did not experience CVD during this follow-up period [6]. In 218 of the initially selected 227 men, baseline samples for PLTP activity assay were available and new-onset T2DM could be ascertained during follow-up.
A. Abbasi et al. / Clinica Chimica Acta 439 (2015) 38–41
2.2. Primary outcome Incident T2DM cases were ascertained if one or more of the following conditions were fulfilled during follow-up: a physician diagnosed T2DM as indicated by self-report, fasting plasma glucose ≥7.0 mmol/l, a random sample plasma glucose ≥ 11.1 mmol/l or initiation of glucose-lowering medication as retrieved from a central pharmacy registry [7]. 2.3. Plasma PLTP activity and other laboratory measurements PLTP activity (in arbitrary units [AU] relative to pool plasma) was measured with a liposome vesicle–HDL system in EDTA-anticoagulated plasma samples that were stored at − 80 °C until assay (intra-assay coefficient of variation b 5%) [3]. All other laboratory assays were also carried out in the baseline samples. Glucose, lipids, lipoproteins and insulin were assayed as described [7]. 2.4. Statistical analysis Statistical analyses were performed using IBM SPSS or R-2.13.1 for Windows (http://cran.r-project.org/). Data are given in mean SD, median (interquartile ranges) or numbers. Between group differences were compared by unpaired t tests, Mann–Whitney U tests and chi-square analysis where appropriate. Univariate relationships were determined by Spearman's rank correlation coefficients (r). Logistic regression was used to test the association of PLTP activity with incident T2DM, since new-onset diabetes has been ascertained periodically in the PREVEND cohort, i.e. at the regular screening visits. Odds ratios (ORs) are given with 95% confidence intervals (CIs). Given the rather low number of subjects with incident T2DM, multivariable adjustment was done for maximally 2 variables together. The additive value of PLTP activity for T2DM prediction was estimated in terms of change in the C-statistic and reclassification (integrated discrimination improvement (IDI)) using the presence of the metabolic syndrome or the Data from the Epidemiological Study on the Insulin Resistance Syndrome (DESIR) risk score (a sex-specific diabetes prediction model, validated in PREVEND) as reference [8]. Two-sided pvalues b 0.05 were considered statistically significant. 3. Results During a follow-up of 9.4 ± 0.7 (mean ± SD) years, 24 participants (11.0%) developed T2DM. Plasma PLTP activity was normally distributed (Kolmogorov–Smirnov test: p = 0.20), and was higher in the individuals who developed T2DM compared to non-convertors (106.4 ± 8.7 vs. 100.3 ± 10.8 AU, p = 0.009; Table 1). In all participants combined, PLTP activity was correlated positively with waist circumference (r = 0.145, p = 0.03), the DESIR risk score (r = 0.160, p = 0.018), plasma insulin (r = 0.216, p = 0.001), insulin resistance (homeostasis model assessment (HOMA-IR)) (r = 0.201, p = 0.003), total cholesterol (r = 0.138, p = 0.042) and triglycerides (r = 0.211, p = 0.002), but not with glucose (r = 0.036, p = 0.599). We observed crude 82% higher odds for T2DM per 1-SD increase in PLTP activity (Table 1). The association of PLTP activity with T2DM remained significant after adjustment for age and either the presence of the metabolic syndrome, individual metabolic syndrome components, total cholesterol, the DESIR risk score, HOMA-IR, the case–control design or albuminuria. Corresponding odds ratios (95% CI) for incident T2DM ranged from 1.64 (1.03–2.66) to 1.87 (1.19–3.01) per 1-SD increase in PLTP activity (Table 2). Addition of PLTP activity to reference models of either the metabolic syndrome or the DESIR prediction model resulted in improvement in the C-statistic or reclassification (IDI). The greatest C-statistic and IDI were achieved when we added PLTP activity along with glucose to the DESIR model (Table 3).
39
Table 1 Baseline clinical and laboratory characteristics and the relationship of plasma phospholipid transfer protein activity with incident type 2 diabetes mellitus. Variablesa
Non-convertor (n = 194)
Incident type 2 diabetes mellitus (n = 24)
Age (years) Incident CVD (number; %) Smoking (number; %) BMI (kg/m2) Waist circumference (cm) Systolic blood pressure (mm Hg) Diastolic blood pressure (mm Hg) Hypertension (number; %) Metabolic syndrome (number; %) DESIR risk score UAE (mg/24 h) Glucose (mmol/l) Insulin (pmol/l)
51.8 ± 12.4 98 (50.5) 90 (46.4) 26.4 ± 3.4 94 ± 10 137 ± 20
57.6 ± 10.1 14 (58.3) 10 (41.7) 28.2 ± 4.5 100 ± 10 144 ± 19
0.03 0.48 0.66 0.02 0.007 0.14
79 ± 9
81 ± 6
0.32
89 (45.6) 46 (23.7)
15 (62.9) 14 (58.3)
0.12 b0.001
2.82 ± 1.21 12.1 (7.5–22.7) 4.84 ± 0.58 47.4 (35.0–73.2) 1.7 (1.2–2.6) 5.88 ± 1.14 1.14 ± 0.36 1.31 (0.90–1.92) 100.3 ± 10.8
3.33 ± 1.09 20.2 (10.2–35.9) 5.47 ± 0.69 73.2 (50.4–115.2) 2.9 (2.0–4.8) 6.24 ± 0.88 1.05 ± 0.34 1.81 (1.45–2.67) 106.4 ± 8.7
0.02 0.03 b0.001 0.001
HOMA-IR Total cholesterol (mmol/l) HDL-cholesterol (mmol/l) Triglycerides (mmol/l) PLTP activity (AU)
p-Value
b0.001 0.13 0.23 0.003 0.009
The metabolic syndrome was defined according to the revised National Cholesterol Education Program's Adult Treatment Panel III report with at least three of the following criteria: 1) waist circumference N102 cm, 2) blood pressure ≥130/85 mm Hg or treatment for hypertension, 3) fasting triglycerides ≥1.7 mmol/l, 4) high-density lipoprotein (HDL) cholesterol ≤1.0 mmol/l, and 5) fasting plasma glucose ≥5.6 mmol/l or treatment for type 2 diabetes mellitus. The DESIR risk score was calculated as a sum of waist circumference, current smoking and hypertension as follows [8]: waist circumference b80 cm: 0; 80–89 cm: 1; 90–99 cm: 2; ≥100 cm: 3; current smoking: 1; hypertension: 1 (total score: 0–5). The absolute risk of type 2 diabetes is calculated as: exp(α + β × linear predictor)/(1 + exp(α + β × linear predictor)), where, α is the calibration intercept, β is the calibration slope, and linear predictor (if men) = −10.45 + 0.72 (if current smoker, otherwise 0) + 0.081 × waist circumference (cm) + 0.50 (if hypertension, otherwise 0). HOMA-IR was calculated as fasting plasma insulin [in mU/l] × fasting plasma glucose [in mmol/l] divided by 22.5. a Data are shown as mean ± SD or as median (interquartile ranges) for continuous variables which were compared by independent t tests or Mann–Whitney U tests where appropriate. Categorical variables are shown as frequencies (percentage), and were compared by χ2 test. Abbreviations: AU, arbitrary units; CVD, cardiovascular disease; BMI, body mass index (the weight in kilogram divided by the square of the height in meters); DESIR, Data from the Epidemiological Study on the Insulin Resistance Syndrome; HDL, high-density lipoproteins; LDL, low-density lipoproteins; HOMA-IR, homeostasis model assessment; PLTP, phospholipid transfer protein; UAE, urinary albumin excretion.
4. Discussion Our prospective study supports the hypothesis that higher plasma PLTP activity is associated with an increased risk for developing T2DM in men. Remarkably, the strength of association of PLTP activity with incident T2DM remained essentially similar when taking account of established diabetes risk factors, HDL cholesterol, triglycerides, total cholesterol, HOMA-IR or albuminuria. Of potential interest, diabetes risk prediction tended to improve when PLTP activity was added to reference models of the metabolic syndrome or the DESIR risk score together with glucose. Our findings suggest that the putative relationship between plasma PLTP activity and glucose homeostasis is at least in part independent from conventional diabetes risk factors and plasma lipid levels. PLTP could influence glucose homeostasis via several yet to be more precisely delineated processes. PLTP deficiency has been recently shown to enhance pancreatic glucose-stimulated insulin secretion and to increase peripheral glucose disposal after lipopolysaccharide administration-induced endotoxemia in mice [9]. Additionally, PLTP gene expression is regulated by liver X receptor-, farnesoid X-activated receptor- and peroxisome proliferator activated receptor-dependent mechanisms, which are also involved in glucose homeostasis [10].
40
A. Abbasi et al. / Clinica Chimica Acta 439 (2015) 38–41
Table 2 Predictive value of plasma phospholipid transfer protein (PLTP) activity for incident type 2 diabetes mellitus by logistic regression analysis. Logistic regression models
OR (95% CI) per 1-SD PLTP activity
p-value
Crude Model 1 Model 2 Model 3 Model 4 Model 5 Model 6 Model 7 Model 8 Model 9 Model 10 Model 11 Model 12
1.82 (1.17–2.91) 1.84 (1.18–2.95) 1.79 (1.13–2.91) 1.87 (1.19–3.01) 1.82 (1.17–2.93) 1.82 (1.15–2.94) 1.86 (1.20–2.99) 1.68 (1.07–2.70) 1.81 (1.16–2.91) 1.74 (1.11–2.80) 1.64 (1.03–2.66) 1.84 (1.18–2.95) 1.84 (1.18–2.96)
0.010 0.008 0.016 0.008 0.010 0.012 0.007 0.028 0.010 0.018 0.039 0.009 0.009
Sources of funding This work was supported by the Netherlands Heart Foundation (2001.005), Dutch Diabetes Research Foundation, Dutch Kidney Foundation, the Netherlands Organization for Scientific Research project (NWO) and the Medical Research Council UK (grant reference no. MC_U106179471). AA is supported by a Rubicon grant from the Netherlands Organization for Scientific Research (NWO). Disclosures
Odds ratios (ORs) with 95% confidence intervals (CI) are expressed per SD increment in PLTP activity Model 1: adjusted for age; model 2: adjusted for age plus the presence of the metabolic syndrome; model 3: adjusted for age plus hypertension criterion; model 4: adjusted for age plus waist circumference criterion; model 5: adjusted for age plus glucose criterion; model 6: adjusted for age plus high density lipoprotein cholesterol criterion; model 7: adjusted for age plus triglyceride criterion; model 8: adjusted for age plus total cholesterol; model 9: adjusted for age plus “Data from the Epidemiological Study on the Insulin Resistance Syndrome” (DESIR) score; model 10: adjusted for age plus homeostasis model assessment; model 11: adjusted for age plus case–control study design; model 12: adjusted for age plus urinary albumin excretion.
PLTP does not only affect key pathways involved in lipoprotein metabolism, including HDL remodeling resulting in enhanced hepatic VLDL secretion and in smaller-sized HDL particles, but may also influence the quality of plasma lipoproteins [1,2]. Furthermore, PLTP exerts several actions at the cellular level, including effects on the nuclear factor-κB pathway in macrophages with functional consequences for vascular biology and inflammation [1,2]. Of additional interest, common genetic variations in PLTP associate with glycemic traits [11]. Remarkably, in contrast to the positive relationship between higher triglycerides, lower HDL cholesterol and higher glucose levels, genetic predisposition to dyslipidemia may associate with a pleiotropic lowering influence on glycemic traits [11]. Limitations of the present study include its small sample size and the lack of power to explore potential interactions. Furthermore, the baseline difference in plasma PLTP activity between men who developed T2DM compared to men who did not was modest (6.1%), possibly limiting its clinical utility. Notably, the cumulative incidence of T2DM in this cohort was quite comparable to that of men without diabetes at baseline in the PREVEND cohort [12]. The relationship of PLTP with diabetes development needs to be replicated, particularly in women and in non-white populations. However, no large cohorts are available with plasma PLTP activity measurements using the original liposome assay. Inherent to an observational study, we could not address the nature of relationship and reverse causation. In conclusion, our results underscore the hypothesis that plasma PLTP activity may serve as a biomarker of incident T2DM, and may provide a rationale to determine possible effects of pharmacological PLTP inhibition on glucose homeostasis.
None of the study sponsors had a role in the study design, data collection, analysis and interpretation, report writing, or the decision to submit the report for publication. All authors declare no conflict of interest, no financial relationships with any organizations that might have an interest in the submitted work in the previous three years, no other relationships or activities that could appear to have influenced the submitted work. All authors declare that there is no duality of interest associated with their contribution to this manuscript. Contribution statement AA and RPD conceived and designed the study. AA performed statistical analysis for the study. AA, RPD and GD-T contributed to materials and analysis tools or to acquisition of data, data analysis and interpretation. AA and RPD wrote the first draft of the manuscript. AA, RPD and GD-T contributed to interpretation of data, critical revision of the manuscript for important intellectual content. AA, RPD and GD-T contributed to final approval of the version to be published. AA and RPD (who supervised the study) had full access to all of the data in the study and were guarantors of the study. Acknowledgments Prof. J.L. Hillege, Department of Cardiology, University of Groningen and University Medical Center Groningen, provided advice in the design of the nested case–control study. The laboratory work of J.J. Duker and J. van der Wal–Haneveld is appreciated. Prof. L.T.W. de Jong-van den Berg and dr. S.T. Visser Department of Social Pharmacy, Pharmacoepidemiology and Pharmacotherapy, Groningen University Institute for Drug Exploration, University of Groningen, University Medical Center Groningen provided the data on pharmacyregistered use of anti-diabetic medications. References [1] Tzotzas T, Desrumaux C, Lagrost L. Plasma phospholipid transfer protein (PLTP): review of an emerging cardiometabolic risk factor. Obes Rev 2009;10:403–11. [2] Dullaart RPF, van Tol A, Dallinga-Thie GM. Phospholipid transfer protein, an emerging cardiometabolic risk marker: is it time to intervene? Atherosclerosis 2013;228: 38–41. [3] Vergeer M, Boekholdt SM, Sandhu MS, et al. Genetic variation at the phospholipid transfer protein locus affects its activity and high-density lipoprotein size and is a novel marker of cardiovascular disease susceptibility. Circulation 2010;122:470–7.
Table 3 Contribution of PLTP activity to change in C-statistic and integration discrimination improvement (IDI).
Metabolic syndrome (yes/no)a Metabolic syndrome plus PLTP activity DESIR scoreb DESIR plus PLTP activity DESIR plus glucose DESIR plus glucose plus PLTP activity a b
C-statistic for models
p-Value for change in C-statistic
IDI for extended models
0.678 (0.575–0.785) 0.756 (0.663–0.848) 0.641 (0.528–0.754) 0.697 (0.594–0.801) 0.771 (0.641–0.879) 0.802 (0.711–0.894)
Reference 0.008 Reference 0.16 0.04 0.006
Reference 0.028 Reference 0.026 0.101 0.128
The presence of the metabolic syndrome was used as reference category. The “Data from the Epidemiological Study on the Insulin Resistance Syndrome” (DESIR) risk score was used as reference category.
p-Value for IDI 0.08 0.03 0.001 b0.001
A. Abbasi et al. / Clinica Chimica Acta 439 (2015) 38–41 [4] Moerland M, Samyn H, van Gent T, et al. Acute elevation of plasma PLTP activity strongly increases pre-existing atherosclerosis. Arterioscler Thromb Vasc Biol 2008;28:1277–82. [5] Schlitt A, Liu J, Yan D, Mondragon-Escorpizo M, Norin AJ, Jiang XC. Antiinflammatory effects of phospholipid transfer protein (PLTP) deficiency in mice. Biochim Biophys Acta 2005;1733:187–91. [6] Borggreve SE, Hillege HL, Dallinga-Thie GM, et al. High plasma cholesteryl ester transfer protein levels may favour reduced incidence of cardiovascular events in men with low triglycerides. Eur Heart J 2007;28:1012–8. [7] Abbasi A, Corpeleijn E, Gansevoort RT, et al. Role of HDL cholesterol and estimates of HDL particle composition in future development of type 2 diabetes in the general population: the PREVEND study. J Clin Endocrinol Metab 2013;98:E1352–9. [8] Balkau B, Lange C, Fezeu L, et al. Predicting diabetes: clinical, biological, and genetic approaches: data from the Epidemiological Study on the Insulin Resistance Syndrome (DESIR). Diabetes Care 2008;31:2056–61.
41
[9] Nguyen AT, Mandard S, Dray C, et al. Lipopolysaccharides-mediated increase in glucose-stimulated insulin secretion: involvement of the GLP-1 pathway. Diabetes 2014;63:471–82. [10] Lakomy D, Rebe C, Sberna AL, et al. Liver X receptor-mediated induction of cholesteryl ester transfer protein expression is selectively impaired in inflammatory macrophages. Arterioscler Thromb Vasc Biol 2009;29:1923–9. [11] Li N, van der Sijde MR, Study LC, et al. Pleiotropic effects of lipid genes on plasma glucose, HbA1c and HOMA-IR levels. Diabetes 2014;63:3149–58. [12] Abbasi A, Corpeleijn E, Meijer E, et al. Sex differences in the association between plasma copeptin and incident type 2 diabetes: the Prevention of Renal and Vascular Endstage Disease (PREVEND) study. Diabetologia 2012;55 [1963-197].