- Email: [email protected]
Gender differences in the relationship between leptin, insulin resistance and the autonomic nervous system
Regulatory Peptides 140 (2007) 37 – 42 www.elsevier.com/locate/regpep
Gender differences in the relationship between leptin, insulin resistance and the autonomic nervous system Daniel E. Flanagan a,⁎, Julian C. Vaile b , Graham W. Petley c , David I. Phillips d , Ian F. Godsland e , Phillip Owens f , Vivienne M. Moore f , Richard A. Cockington f , Jeffrey S. Robinson f a Department of Endocrinology, Derriford Hospital, Plymouth, UK Department of Cardiovascular Medicine, Queen Elizabeth Medical Centre, University of Birmingham, UK Department of Medical Physics, University of Southampton, Southampton General Hospital, Southampton, UK d MRC Environmental Epidemiology Unit, University of Southampton, UK e Endocrinology and Metabolic Medicine, Imperial College School of Medicine, London, UK f Department of Obstetrics and Gynaecology, University of Adelaide, South Australia, Australia b
c
Received 17 March 2006; received in revised form 8 November 2006; accepted 10 November 2006 Available online 21 December 2006
Abstract Objectives: Leptin, an important hormonal regulator of body weight, has been shown to stimulate the sympathetic nervous system (SNS) in vitro although the physiological relevance remains unclear. Increased SNS activity has been implicated in the pathogenesis of insulin resistance and an increased cardiovascular risk. We have therefore investigated the relationship between leptin, insulin resistance and cardiac autonomic activity in healthy young adults. 130 healthy men and women age 20.9 years were studied. Insulin sensitivity was assessed using the IVGTT and minimal model with simultaneous measures of leptin. Cardiac autonomic activity was assessed using spectral analysis of heart rate variability. Results: Women showed significantly higher fasting leptin, heart rate and cardiac sympathetic activity, and lower insulin sensitivity. Men showed inverse correlations between insulin resistance and heart rate, and between insulin resistance and cardiac sympatho-vagal ratio. Women, in contrast, showed no SNS relationship with insulin resistance, but rather an inverse correlation between leptin and the sympatho-vagal ratio, suggesting that leptin in women is associated with SNS activity. The correlation remained significant after adjustment for BMI and waist-to-hip ratio (beta = −0.33 and p = 0.008). Conclusion: Insulin resistance and SNS activity appear to be linked, although the relationship showed marked gender differences, and the direction of causality was unclear from this cross-sectional study. Leptin appears to exert a greater effect on the SNS in women, possibly because of their greater fat mass. © 2006 Elsevier B.V. All rights reserved. Keywords: Leptin; Heart rate; Heart rate variability; Insulin resistance; Sympathetic nervous system
1. Introduction The prevalence of obesity in industrialized nations continues to rise. Adiposity is associated with an increased risk of cardiometabolic syndrome and insulin resistance appears to be an important link between the two. Our understanding of the mechanisms controlling body weight and energy homeostasis ⁎ Corresponding author. Department of Endocrinology, Level 10, Derriford Hospital, Plymouth PL6 8DH, Devon, UK. Tel.: +44 01752 517578; fax: +44 01752 517617. E-mail address: [email protected] (D.E. Flanagan). 0167-0115/$ - see front matter © 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.regpep.2006.11.009
continues to grow, but the mechanisms whereby overweight is translated into the features of the metabolic syndrome remain unclear. Leptin is a crucial hormone in the regulation of body weight. It is produced by adipose tissue and acts centrally by decreasing appetite and increasing energy expenditure [1]. There is a strong positive relationship between leptin and insulin resistance, and leptin behaves as a component of the metabolic syndrome [2]. Moreover, exogenous leptin has been shown experimentally to stimulate sympathetic nervous (SNS) system activity in animal models [3]. Increased SNS activity has been implicated in the pathogenesis of insulin resistance and increased cardiovascular risk [4]. In the present study, we used
38
D.E. Flanagan et al. / Regulatory Peptides 140 (2007) 37–42
cardiac autonomic activity as a surrogate measure of SNS activity in young healthy subjects with normal glucose tolerance, measured by spectral analysis of heart rate variability. We then explored the inter-relationships between cardiac autonomic activity, insulin sensitivity/secretion and leptin concentrations. 2. Methods The study sample comprised 68 males (mean age 20.9 SEM ± 0.03 years) and 62 females (mean age 20.9 SEM ± 0.04 years) drawn from an existing prospective cohort of young adults born in Adelaide, South Australia [5]. Subjects suffering from diabetes or other current illness were excluded. Investigation of the relationship between leptin and insulin resistance was an a priori part of the study design. The protocol was approved by the Human Ethics Committee of the Women's and Children's Hospital (Adelaide, South Australia). Informed written consent was obtained from each of the subjects. Subjects were encouraged to consume more than 200 g/day carbohydrate for three days. They were asked to fast overnight and to refrain from smoking and alcohol overnight, before attending the research department between 08.00 and 09.00. Medical history, smoking habits, alcohol consumption and level of physical activity were recorded by supervised questionnaire. Anthropometric measures were recorded as previously described [6]. An electrocardiogram (ECG) was performed on each subject with a portable analogue tape recorder (Oxford Instruments Co. Ltd., Osney Mead, Oxford) using bipolar skin electrodes applied to the chest wall. Subjects lay supine on a couch in a quiet room, and a 30 minute rest period was allowed to enable heart rate, blood pressure and ventilation to stabilise before the start of each recording. A 15 minute undisturbed ECG recording was obtained. Subjects underwent a 15-point frequently sampled intravenous glucose tolerance test (IVGTT) [7,8]. Plasma samples were analysed for glucose by the hexokinase method and insulin by two site immunometric assays [9]. The within-assay coefficient of variation of the insulin measurements was less than 10%. Leptin was measured in plasma by ELISA (10-23100 Diagnostics Systems Laboratories Inc, Webster TX, USA) according to the manufacturer's instructions. The intra-assay coefficients of variation for repeated measurement of leptin samples at concentrations of 2.0 and 15.3 ng/ml were 7.1% and 1.8% respectively. 2.1. Data analysis Insulin sensitivity (Si) was determined from the IVGTT glucose and insulin profiles using the minimal model of glucose disappearance [7]. The ECG tape cassettes were analysed as previously described [6]. Frequency domain analysis was performed to determine the power of the underlying component oscillations. Power spectral analysis was performed using the Burg Algorithm [10] with a model order between 8 and 12 [11]. The power of each underlying frequency was quantified by decomposing the total variability signal with the method of Zetterberg [12]. This enabled the determination of power at the
two major peaks in the heart rate variability spectrum: low frequency (LF) power (arising between 0.05 to 0.15 Hz) and high frequency (HF) power (0.15 to 0.40 Hz). The high frequency component is a measure of respiratory sinus arrhythmia, mediated almost entirely by the parasympathetic vagus nerve [13]. Low frequency power is thought to be a resonance phenomenon due to delay in the sympathetic feedback loops of the baroreflex system and contains both sympathetic and parasympathetic contributions. Hence, the ratio of HF to LF is considered to represent the balance between sympathetic and parasympathetic contributions to heart rate variability. Because total power varies greatly between individual subjects, it was expressed as normalized units, calculated by dividing the absolute power of a given component (area under the component curve) by the total variance minus the DC component [13]. 2.2. Statistical methods Results are presented as mean ± standard error of the mean (SEM). Where appropriate, the data was log-transformed. Paired sample t tests were used to localize effects found in the initial set of ANOVA. The data were analysed using multiple linear regression, and all analyses were undertaken using continuous variables. 3. Results Table 1 compares the mean age, BMI, measures of body fat distribution, metabolic variables and measures of heart rate variability in men and women. The genders were of similar age Table 1 Metabolic and anthropometric indices for 68 men and 62 women studied Men Age (years) BMI (kg/m2) Waist/hip ratio Biceps skinfold thickness (mm) Triceps skinfold thickness (mm) Supra-iliac skinfold thickness (mm) Subscapular skinfold thickness (mm) Systolic bp (mm Hg) Diastolic bp (mm Hg) Fasting insulin (pmol/l) Si (104 min− 1/pmol l) Fasting leptin (ng/ml) Heart rate (bpm) High frequency (normalized units) Low frequency (normalized units) High/low ratio
Women
p value p value adjusted for adiposity
20.9 (0.22) 20.9 (0.33) 0.494 23.9 (3.7) 23.2 (5.3) 0.184 0.82 (0.06) 0.70 (0.04) <0.001 4.9 (2.7) 8.3 (4.5) <0.001 9.2 (4.5)
17.0 (6.8)
<0.001
11.2 (7.5)
13.2 (7.9)
0.025
10.7 (5.7)
13.8 (7.9)
0.006
124 (9) 70 (8) 40.6 (24.2) 5.74 (3.2) 3.0 (4.8) 55 (7) 41.4 (12.9)
114 (8) <0.001 <0.001 69 (7) 0.223 0.091 51.6 (28.7) 0.003 0.601 4.38 (2.6) 0.009 0.844 13.0 (11.3) <0.001 0.601 59 (8) 0.005 0.009 49.5 (13.9) 0.001 0.001
33.7 (14.0) 26.7 (12.8) 1.6 (1.1)
2.5 (1.8)
0.004
0.005
<0.001 <0.001
Data displayed as mean (standard deviation). Heart rate measures were available for 125 of these subjects (67 men, 58 women). Sum of the four skinfolds is used as a comparison of adiposity between genders.
D.E. Flanagan et al. / Regulatory Peptides 140 (2007) 37–42
and BMI, but women showed significantly greater skinfold thickness at each site measured. Skinfold thickness is highly correlated with adiposity (% body fat). The waist-to-hip ratio was lower in the females, reflecting a likely lower visceral fat mass. Women, however, also had higher fasting insulin, higher fasting leptin and lower insulin sensitivity than men. Gender differences in insulin, insulin sensitivity and leptin remained significant when adjusted for BMI but not when the sum of the four skinfold measures is used as the adiposity measure. Women showed higher resting heart rate and a lower systolic blood pressure. Heart rate variability analysis showed a greater contribution of parasympathetic cardiac activity and less sympathetic activity in the women compared with men, reflected in a higher high frequency to low frequency ratio. Accordingly, data for men and women were analysed separately. The gender differences in spectral analysis could not be explained by differences in anthropometric measures, or by levels of physical activity. 3.1. Leptin and insulin Fig. 1a shows a strong relationship between BMI and fasting leptin in both genders. Leptin levels were generally higher in the
39
women than the men for the same BMI, consistent with their greater adiposity. In men, only the supra-iliac skinfold measure remained a significant predictor of leptin in a stepwise regression. Both the supra-iliac skinfold and waist/hip ratio remained significant in women. A similar pattern emerged for the association between leptin and fasting insulin (Fig. 1b). The relationships between leptin/insulin and insulin/body fat were independent of each other in both genders. Thus, leptin rose with body fat, but also independently with insulin. Fig. 2 shows a significant rise in leptin among the women following bolus iv glucose (13.0 ± 1.4 to 15.4 ± 1.7 ng/ml and p = 0.017), but none in the men. 3.2. Leptin and cardiac autonomic activity Increased fasting leptin was associated with increasing heart rate in women, but not in men (Fig. 3). Using a stepwise linear regression adjustment for BMI and other measures of body fat (waist hip ratio, each skinfold measure and the sum of skinfolds) did not alter this relationship. Table 2 again shows marked gender differences in the correlations between fasting plasma leptin, the measures of heart rate variability and blood
Fig. 1. (a) Relationship between fasting leptin and BMI for 68 men (r = 0.73 and p < 0.001) and 62 women (r = 0.72 and p < 0.001). (b) Relationship between fasting insulin and fasting leptin for 68 men (r = 0.83 and p < 0.001) and 62 women (r = 0.48 and p < 0.001).
40
D.E. Flanagan et al. / Regulatory Peptides 140 (2007) 37–42 Table 2 Correlation between fasting plasma leptin, heart rate variability measures and blood pressure of 68 men and 62 women Men
HR (bpm) SBP (mm Hg) DBP (mm Hg) HF (nU) LF (nU) HF/LF
Fig. 2. Plasma leptin with time following intravenous glucose. Data displayed as mean (se).
pressure. In women, leptin correlates positively with LF power and negatively with the HF/LF ratio, suggesting that leptin is associated with increased cardiac sympathetic and decreased cardiac vagal tone (these relationships remain highly significant allowing for any of the measures of adiposity). Despite this
Women
r
p
r
p
0.10 0.19 0.24 0.06 0.02 0.01
0.419 0.132 0.056 0.617 0.865 0.919
0.27 0.09 0.17 − 0.15 0.42 − 0.33
0.032 0.507 0.200 0.243 0.001 0.008
relationship, there was no association between leptin and blood pressure. While the association between heart rate variability and leptin was not seen in men, a relationship between heart rate and insulin sensitivity was only seen in men [6]. In a stepwise multiple regression analysis in the men, with HF/LF ratio as the outcome variable, only Si (beta = 0.438 and p < 0.001) and supra-iliac skinfold thickness (beta = 0.33 and p = 0.013) remained significant among all metabolic variables and measures of body anthropometry. Only leptin remained significant (beta = − 0.33 and p = 0.008) in the corresponding regression in women. The relationships described above were not confounded by levels of physical activity, alcohol consumption, smoking, stage of the menstrual cycle or hormonal contraceptive use. 4. Discussion
Fig. 3. Relationship between fasting leptin and resting heart rate for 68 men (r = 0.10 and p = 0.419) and 62 women (r = 0.27 and p = 0.032).
Numerous studies have reported gender differences in the physiology of leptin, but it remains unclear whether the dichotomy reflects differences in total body fat, or in body fat distribution [14–16]. Previous work has also indicated an association between leptin and the SNS [17]. Our data confirm this, but find the association to be closer in women than in men. Although there was no gender difference in BMI, women were considerably more adipose than men and their circulating leptin concentrations were four fold higher. BMI is a poor index of fat mass/distribution, and supra-iliac skinfold thickness – a proxy for % body fat – was the strongest predictor of fasting leptin in both sexes. Fasting insulin was a stronger predictor of fasting leptin than insulin sensitivity measured by IVGTT, and in women the waist–hip ratio, an index of central fat distribution, was an additional predictor of fasting leptin. Although fasting insulin is positively associated with fasting leptin, it is not clear which is causal, or indeed whether some cocorrelate is responsible. The rise in leptin following an intravenous bolus of glucose might suggest that insulin drives leptin secretion, a view supported by the rise in circulating leptin reported within five days of starting insulin therapy in patients with newly diagnosed type 1 diabetes [18]. Hyperinsulinemic euglycemic clamp studies, however, show that it is the rate of glucose uptake into adipose tissue rather than the prevailing insulin levels that determine leptin secretion [19]. In vitro experiments support this view [20], though plasma glucose was not an independent predictor of fasting leptin in the present study.
D.E. Flanagan et al. / Regulatory Peptides 140 (2007) 37–42
Resting heart rate is a highly variable measure. Although we carried out this study under standardized conditions, using a relatively long resting ECG (15 min), it was impossible to control completely for anxiety. Although we found strong relationships between heart rate, measures of heart rate variability, leptin and insulin resistance, they are still likely to be underestimates because of measurement ‘noise’. The power spectral analysis of heart rate variability suggests that the relationship between heart rate and metabolic variables is mediated through alterations in cardiac autonomic tone. Although cardiac automaticity is intrinsic, the resting pulse is largely under the control of the autonomic nervous system. Power spectral analysis enables quantitative evaluation of beatto-beat cardiovascular control, and distinguishes the two major components of the spectrum of heart rate variability, namely the high and low frequency components [21]. The high frequency component is centered at the respiratory frequency and is a measure of respiratory sinus arrhythmia. Since respiratory sinus arrhythmia is mediated almost entirely by the vagus nerve, HF power is taken as an index of cardiac vagal control [13]. Low frequency power is thought to be a resonance phenomenon due to delay in the sympathetic feedback loops of the baroreflex system and contains both sympathetic and parasympathetic contributions. Hence, the ratio of HF to LF is considered to represent the balance between vagal and sympathetic contributions to heart rate variability [13,22]. We have previously published evidence of marked gender differences in the relationship between Si and heart rate. Men, but not women, showed a strong relationship between resting heart rate and Si — an increasing heart rate being associated with increasing insulin resistance, independent of BMI. Increasing insulin resistance was also associated with increased sympathetic and decreased parasympathetic cardiac tone [6]. The present data suggest that fasting leptin is more closely related to resting heart rate in young women than young men. Fasting leptin was also correlated with the low frequency component of heart rate variability and the high to low frequency ratio, suggesting that increased circulating leptin is associated with increased cardiac sympathetic and decreased parasympathetic tone. The techniques available for assessing human autonomic activity are often technically difficult or imprecise. We have chosen to measure cardiac autonomic activity because it is non-invasive and technically easy to perform. Prospective studies have shown that measures of heart rate variability are predictive of cardiovascular mortality and morbidity [23]. Previous work suggests that the SNS is an important determinant of leptin secretion. It is known to be important in mobilising fatty acids from adipose tissue [24], and is an important regulator of leptin secretion from adipose tissue. Cold exposure or fasting, both of which raise SNS activity in fat, leads to decreased leptin production [25]. Conversely, sympathetic blockade may lead to increased leptin production [26]. The work by Haynes et al. [3,27] would support the view that leptin is acting to increase sympathetic nervous system activity, this may reflect differing autonomic effects in different tissues. Our results would support the latter view. SNS activity is
41
reduced in rodents with leptin deficiency or leptin receptor defects, contributing to decreased energy expenditure [27]. Obesity is associated with increased measures of SNS activity in humans, and animal work has shown that pharmacological doses of leptin increase SNS activity and blood pressure [28]. One previous study has examined the relationship between heart rate variability and plasma leptin in an older population of men [15]. In this study, mean age of the subjects was 42.5 years and, as would be expected, the subjects were more insulin resistant with higher fasting insulin and leptin concentrations. Interestingly, both HOMA measures of insulin resistance and leptin were predictive of cardiac autonomic activity. This may reflect changing relationships between leptin and the insulin resistance syndrome with time, and illustrates the difficulty in separating the complex relationships between the different systems. We could find no relationship between leptin and blood pressure in our young population. Our observational data would support the hypothesis that leptin may be contributing to the gender differences in the relationship between adiposity and cardiovascular risk — a view that is supported by a number of other groups [29,30]. In summary, we have shown major gender differences in the relationships between leptin, cardiac autonomic activity and insulin/insulin resistance. Importantly, the study was performed in a population of young adults, before features of the insulin resistance syndrome had become established. In women, but not in men, increased leptin is associated with a shift in cardiac autonomic activity towards increased sympathetic tone and greater cardiometabolic risk. The findings may be related to the intrinsically higher insulin resistance recently reported in women. Acknowledgements The authors would like to thank Sister M. Logan, Ms M. Rourke, Ms L. Raggett for assistance with the fieldwork. This study was generously supported by the Medical Research Council, Diabetes UK, the Wessex Medical Trust and the Women's and Children's Hospital Foundation. VMM was supported by the PHRDC, National Health and Medical Research Council of Australia. IFG is supported by the Heart Disease and Diabetes Research Trust. References [1] Zhang Y, Proenca R, Maffei M, Barone M, Leopold L, Friedman JM. Positional cloning of the mouse obese gene and its human homologue. Nature 1994;372:425–32. [2] Leyva F, Godsland IF, Ghatei M, Proudler AJ, Aldis S, Walton C, Bloom S, Stevenson JC. Hyperleptinemia as a component of a metabolic syndrome of cardiovascular risk. Arterioscler Thromb Vasc Biol 1998;18:928–33. [3] Haynes WG, Morgan DA, Walsh SA, Mark AL, Sivitz WI. Receptormediated regional sympathetic nerve activation by leptin. J Clin Invest 1997;100:270–8. [4] Facchini FS, Stoohs RA, Reaven GM. Enhanced sympathetic nervous system activity. The linchpin between insulin resistance, hyperinsulinemia, and heart rate. Am J Hypertens 1996;9:1013–7. [5] Flanagan DE, Moore VM, Godsland IF, Cockington RA, Robinson JS, Phillips DI. Fetal growth and the physiological control of glucose tolerance in adults: a minimal model analysis. Am J Physiol 2000;278:E700–6.
42
D.E. Flanagan et al. / Regulatory Peptides 140 (2007) 37–42
[6] Flanagan DE, Vaile JC, Petley GW, Moore VM, Godsland IF, Cockington RA, Robinson JS, Phillips DIW. The autonomic control of heart rate and insulin resistance in young adults. J Clin Endocrinol Metab 1999;84:1263–7. [7] Bergman RN, Ider Y, Bowden C, Cobelli C. Quantitative estimation of insulin sensitivity. Am J Physiol 1979;236:E667–77. [8] Walton C, Godsland IF, Proudler AJ, Felton C, Wynn V. Evaluation of four mathematical models of glucose and insulin dynamics with analysis of effects of age and obesity. Am J Physiol 1992;262:E755–62. [9] Alpha B, Cox L, Crowther N, Clark PM, Hales CN. Sensitive amplified immunoenzymometric assays (IEMA) for human insulin and intact proinsulin. Eur J Clin Chem Clin Biochem 1992;30:27–32. [10] Burg JP. A new analysis technique for time series data. Erschede Netherlands: NATO Advanced Study Institute on signal processing with emphasis on underwater acoustics; 1968. [11] Akaike H. Statistical predictor identification. Ann Inst Stat Math 1970;22:203–17. [12] Zetterberg LH. Estimation of parameters for a linear difference equation with application to EEG analysis. Math Biosci 1969;5:227–75. [13] Malliani A, Pagani M, Lombardi F, Cerutti S. Cardiovascular neural regulation explored in the frequency domain. Circulation 1991;84:482–92. [14] Ostlund Jr RE, Yang JW, Klein S, Gingerich R. Relation between plasma leptin concentration and body fat, gender, diet, age, and metabolic covariates. J Clin Endocrinol Metab 1996;81:3909–13. [15] Paolisso G, Rizzo MR, Mone CM, Tagliamonte MR, Gambardella A, Riondino M, Carella C, Varricchio M, D'Onofrio F. Plasma sex hormones are significantly associated with plasma leptin concentration in healthy subjects. Clin Endocrinol 1998;48:291–7. [16] Liuzzi A, Savia G, Tagliaferri M, Lucantoni R, Berselli ME, Petroni ML, De Medici C, Viberti GC. Serum leptin concentration in moderate and severe obesity: relationship with clinical, anthropometric and metabolic factors. Int J Obes Relat Metab Disord 1999;23:1066–73. [17] Rayner DV, Trayhurn P. Regulation of leptin production: sympathetic nervous system interactions. J Mol Med 2001;79:8–20. [18] Soliman AT, Omar M, Assem HM, Ibrahim SN, Rizk MM, El Matary W, El Alaily RK. Serum leptin concentrations in children with type 1 diabetes mellitus: relationship to body mass index insulin dose and glycemic control. Metabolism 2002;51:292–6. [19] Wellhoener P, Fruehwald-Schultes B, Kern W, Dantz D, Kerner W, Born J, Fehm HL, Peters A. Glucose metabolism rather than insulin is a main
[20] [21]
[22]
[23]
[24]
[25]
[26]
[27] [28]
[29]
[30]
determinant of leptin secretion in humans. J Clin Endocrinol Metab 2000;85:1267–71. Levy JR, Stevens W. The effects of insulin, glucose, and pyruvate on the kinetics of leptin secretion. Endocrinology 2001;142:3558–62. Anonymous. Heart rate variability. Standards of measurement, physiological interpretation, and clinical use. Task Force of the European Society of Cardiology and the North American Society of Pacing and Electrophysiology. Eur Heart J 1996;17:354–81. Akselrod S, Gordon D, Ubel FA, Shannon DC, Barger AC, Cohen RJ. Power spectrum analysis of heart rate fluctuations: a quantitative probe of beat to beat cardiovascular control. Science 1981;213:220–2. La Rovere MT, Bigger Jr JT, Marcus FI, Mortara A, Schwartz PJ, ATRAMI (Autonomic Tone and Reflexes After Myocardial Infarction) Investigators. Baroreflex sensitivity and heart-rate variability in prediction of total cardiac mortality after myocardial infarction. Lancet 1998;351:478–84. Millet L, Barbe P, Lafontan M, Berlan M, Galitzky J. Catecholamine effects on lipolysis and blood flow in human abdominal and femoral adipose tissue. J Appl Physiol 1998;85:181–8. Trayhurn P, Duncan JS, Rayner DV. Acute cold-induced suppression of ob (obese) gene expression in white adipose tissue of mice: mediation by the sympathetic system. Biochem J 1995;311:729–33. Evans BA, Agar L, Summers RJ. The role of the sympathetic nervous system in the regulation of leptin synthesis in C57BL/6 mice. FEBS Lett 1999;444:149–54. Haynes WG. Interaction between leptin and sympathetic nervous system in hypertension. Curr Hyperten Rep 2000;2:311–8. Satoh N, Ogawa Y, Katsuura G, Numata Y, Tsuji T, Hayase M, Ebihara K, Masuzaki H, Hosoda K, Yoshimasa Y, Nakao K. Sympathetic activation of leptin via the ventromedial hypothalamus: leptin-induced increase in catecholamine secretion. Diabetes 1999;48:1787–93. Lyoussi B, Ragala MA, Mguil M, Chraibi A, Israili ZH. Gender specific leptinaemia and its relationship with some components of the metabolic syndrome in Moroccans. Clin Exp Hypertens 2005;27:377–94. Couillard C, Mauriege P, Prud’homme D, Nadeau A, Tremblay A, Bouchard C, Despres JP. Plasma leptin concentrations: gender differences and associations with the metabolic risk factors for cardiovascular disease. Diabetologia 1997;40:1178–84.
Gender differences in the relationship between leptin, insulin resistance and the autonomic nervous system Daniel E. Flanagan a,⁎, Julian C. Vaile b , Graham W. Petley c , David I. Phillips d , Ian F. Godsland e , Phillip Owens f , Vivienne M. Moore f , Richard A. Cockington f , Jeffrey S. Robinson f a Department of Endocrinology, Derriford Hospital, Plymouth, UK Department of Cardiovascular Medicine, Queen Elizabeth Medical Centre, University of Birmingham, UK Department of Medical Physics, University of Southampton, Southampton General Hospital, Southampton, UK d MRC Environmental Epidemiology Unit, University of Southampton, UK e Endocrinology and Metabolic Medicine, Imperial College School of Medicine, London, UK f Department of Obstetrics and Gynaecology, University of Adelaide, South Australia, Australia b
c
Received 17 March 2006; received in revised form 8 November 2006; accepted 10 November 2006 Available online 21 December 2006
Abstract Objectives: Leptin, an important hormonal regulator of body weight, has been shown to stimulate the sympathetic nervous system (SNS) in vitro although the physiological relevance remains unclear. Increased SNS activity has been implicated in the pathogenesis of insulin resistance and an increased cardiovascular risk. We have therefore investigated the relationship between leptin, insulin resistance and cardiac autonomic activity in healthy young adults. 130 healthy men and women age 20.9 years were studied. Insulin sensitivity was assessed using the IVGTT and minimal model with simultaneous measures of leptin. Cardiac autonomic activity was assessed using spectral analysis of heart rate variability. Results: Women showed significantly higher fasting leptin, heart rate and cardiac sympathetic activity, and lower insulin sensitivity. Men showed inverse correlations between insulin resistance and heart rate, and between insulin resistance and cardiac sympatho-vagal ratio. Women, in contrast, showed no SNS relationship with insulin resistance, but rather an inverse correlation between leptin and the sympatho-vagal ratio, suggesting that leptin in women is associated with SNS activity. The correlation remained significant after adjustment for BMI and waist-to-hip ratio (beta = −0.33 and p = 0.008). Conclusion: Insulin resistance and SNS activity appear to be linked, although the relationship showed marked gender differences, and the direction of causality was unclear from this cross-sectional study. Leptin appears to exert a greater effect on the SNS in women, possibly because of their greater fat mass. © 2006 Elsevier B.V. All rights reserved. Keywords: Leptin; Heart rate; Heart rate variability; Insulin resistance; Sympathetic nervous system
1. Introduction The prevalence of obesity in industrialized nations continues to rise. Adiposity is associated with an increased risk of cardiometabolic syndrome and insulin resistance appears to be an important link between the two. Our understanding of the mechanisms controlling body weight and energy homeostasis ⁎ Corresponding author. Department of Endocrinology, Level 10, Derriford Hospital, Plymouth PL6 8DH, Devon, UK. Tel.: +44 01752 517578; fax: +44 01752 517617. E-mail address: [email protected] (D.E. Flanagan). 0167-0115/$ - see front matter © 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.regpep.2006.11.009
continues to grow, but the mechanisms whereby overweight is translated into the features of the metabolic syndrome remain unclear. Leptin is a crucial hormone in the regulation of body weight. It is produced by adipose tissue and acts centrally by decreasing appetite and increasing energy expenditure [1]. There is a strong positive relationship between leptin and insulin resistance, and leptin behaves as a component of the metabolic syndrome [2]. Moreover, exogenous leptin has been shown experimentally to stimulate sympathetic nervous (SNS) system activity in animal models [3]. Increased SNS activity has been implicated in the pathogenesis of insulin resistance and increased cardiovascular risk [4]. In the present study, we used
38
D.E. Flanagan et al. / Regulatory Peptides 140 (2007) 37–42
cardiac autonomic activity as a surrogate measure of SNS activity in young healthy subjects with normal glucose tolerance, measured by spectral analysis of heart rate variability. We then explored the inter-relationships between cardiac autonomic activity, insulin sensitivity/secretion and leptin concentrations. 2. Methods The study sample comprised 68 males (mean age 20.9 SEM ± 0.03 years) and 62 females (mean age 20.9 SEM ± 0.04 years) drawn from an existing prospective cohort of young adults born in Adelaide, South Australia [5]. Subjects suffering from diabetes or other current illness were excluded. Investigation of the relationship between leptin and insulin resistance was an a priori part of the study design. The protocol was approved by the Human Ethics Committee of the Women's and Children's Hospital (Adelaide, South Australia). Informed written consent was obtained from each of the subjects. Subjects were encouraged to consume more than 200 g/day carbohydrate for three days. They were asked to fast overnight and to refrain from smoking and alcohol overnight, before attending the research department between 08.00 and 09.00. Medical history, smoking habits, alcohol consumption and level of physical activity were recorded by supervised questionnaire. Anthropometric measures were recorded as previously described [6]. An electrocardiogram (ECG) was performed on each subject with a portable analogue tape recorder (Oxford Instruments Co. Ltd., Osney Mead, Oxford) using bipolar skin electrodes applied to the chest wall. Subjects lay supine on a couch in a quiet room, and a 30 minute rest period was allowed to enable heart rate, blood pressure and ventilation to stabilise before the start of each recording. A 15 minute undisturbed ECG recording was obtained. Subjects underwent a 15-point frequently sampled intravenous glucose tolerance test (IVGTT) [7,8]. Plasma samples were analysed for glucose by the hexokinase method and insulin by two site immunometric assays [9]. The within-assay coefficient of variation of the insulin measurements was less than 10%. Leptin was measured in plasma by ELISA (10-23100 Diagnostics Systems Laboratories Inc, Webster TX, USA) according to the manufacturer's instructions. The intra-assay coefficients of variation for repeated measurement of leptin samples at concentrations of 2.0 and 15.3 ng/ml were 7.1% and 1.8% respectively. 2.1. Data analysis Insulin sensitivity (Si) was determined from the IVGTT glucose and insulin profiles using the minimal model of glucose disappearance [7]. The ECG tape cassettes were analysed as previously described [6]. Frequency domain analysis was performed to determine the power of the underlying component oscillations. Power spectral analysis was performed using the Burg Algorithm [10] with a model order between 8 and 12 [11]. The power of each underlying frequency was quantified by decomposing the total variability signal with the method of Zetterberg [12]. This enabled the determination of power at the
two major peaks in the heart rate variability spectrum: low frequency (LF) power (arising between 0.05 to 0.15 Hz) and high frequency (HF) power (0.15 to 0.40 Hz). The high frequency component is a measure of respiratory sinus arrhythmia, mediated almost entirely by the parasympathetic vagus nerve [13]. Low frequency power is thought to be a resonance phenomenon due to delay in the sympathetic feedback loops of the baroreflex system and contains both sympathetic and parasympathetic contributions. Hence, the ratio of HF to LF is considered to represent the balance between sympathetic and parasympathetic contributions to heart rate variability. Because total power varies greatly between individual subjects, it was expressed as normalized units, calculated by dividing the absolute power of a given component (area under the component curve) by the total variance minus the DC component [13]. 2.2. Statistical methods Results are presented as mean ± standard error of the mean (SEM). Where appropriate, the data was log-transformed. Paired sample t tests were used to localize effects found in the initial set of ANOVA. The data were analysed using multiple linear regression, and all analyses were undertaken using continuous variables. 3. Results Table 1 compares the mean age, BMI, measures of body fat distribution, metabolic variables and measures of heart rate variability in men and women. The genders were of similar age Table 1 Metabolic and anthropometric indices for 68 men and 62 women studied Men Age (years) BMI (kg/m2) Waist/hip ratio Biceps skinfold thickness (mm) Triceps skinfold thickness (mm) Supra-iliac skinfold thickness (mm) Subscapular skinfold thickness (mm) Systolic bp (mm Hg) Diastolic bp (mm Hg) Fasting insulin (pmol/l) Si (104 min− 1/pmol l) Fasting leptin (ng/ml) Heart rate (bpm) High frequency (normalized units) Low frequency (normalized units) High/low ratio
Women
p value p value adjusted for adiposity
20.9 (0.22) 20.9 (0.33) 0.494 23.9 (3.7) 23.2 (5.3) 0.184 0.82 (0.06) 0.70 (0.04) <0.001 4.9 (2.7) 8.3 (4.5) <0.001 9.2 (4.5)
17.0 (6.8)
<0.001
11.2 (7.5)
13.2 (7.9)
0.025
10.7 (5.7)
13.8 (7.9)
0.006
124 (9) 70 (8) 40.6 (24.2) 5.74 (3.2) 3.0 (4.8) 55 (7) 41.4 (12.9)
114 (8) <0.001 <0.001 69 (7) 0.223 0.091 51.6 (28.7) 0.003 0.601 4.38 (2.6) 0.009 0.844 13.0 (11.3) <0.001 0.601 59 (8) 0.005 0.009 49.5 (13.9) 0.001 0.001
33.7 (14.0) 26.7 (12.8) 1.6 (1.1)
2.5 (1.8)
0.004
0.005
<0.001 <0.001
Data displayed as mean (standard deviation). Heart rate measures were available for 125 of these subjects (67 men, 58 women). Sum of the four skinfolds is used as a comparison of adiposity between genders.
D.E. Flanagan et al. / Regulatory Peptides 140 (2007) 37–42
and BMI, but women showed significantly greater skinfold thickness at each site measured. Skinfold thickness is highly correlated with adiposity (% body fat). The waist-to-hip ratio was lower in the females, reflecting a likely lower visceral fat mass. Women, however, also had higher fasting insulin, higher fasting leptin and lower insulin sensitivity than men. Gender differences in insulin, insulin sensitivity and leptin remained significant when adjusted for BMI but not when the sum of the four skinfold measures is used as the adiposity measure. Women showed higher resting heart rate and a lower systolic blood pressure. Heart rate variability analysis showed a greater contribution of parasympathetic cardiac activity and less sympathetic activity in the women compared with men, reflected in a higher high frequency to low frequency ratio. Accordingly, data for men and women were analysed separately. The gender differences in spectral analysis could not be explained by differences in anthropometric measures, or by levels of physical activity. 3.1. Leptin and insulin Fig. 1a shows a strong relationship between BMI and fasting leptin in both genders. Leptin levels were generally higher in the
39
women than the men for the same BMI, consistent with their greater adiposity. In men, only the supra-iliac skinfold measure remained a significant predictor of leptin in a stepwise regression. Both the supra-iliac skinfold and waist/hip ratio remained significant in women. A similar pattern emerged for the association between leptin and fasting insulin (Fig. 1b). The relationships between leptin/insulin and insulin/body fat were independent of each other in both genders. Thus, leptin rose with body fat, but also independently with insulin. Fig. 2 shows a significant rise in leptin among the women following bolus iv glucose (13.0 ± 1.4 to 15.4 ± 1.7 ng/ml and p = 0.017), but none in the men. 3.2. Leptin and cardiac autonomic activity Increased fasting leptin was associated with increasing heart rate in women, but not in men (Fig. 3). Using a stepwise linear regression adjustment for BMI and other measures of body fat (waist hip ratio, each skinfold measure and the sum of skinfolds) did not alter this relationship. Table 2 again shows marked gender differences in the correlations between fasting plasma leptin, the measures of heart rate variability and blood
Fig. 1. (a) Relationship between fasting leptin and BMI for 68 men (r = 0.73 and p < 0.001) and 62 women (r = 0.72 and p < 0.001). (b) Relationship between fasting insulin and fasting leptin for 68 men (r = 0.83 and p < 0.001) and 62 women (r = 0.48 and p < 0.001).
40
D.E. Flanagan et al. / Regulatory Peptides 140 (2007) 37–42 Table 2 Correlation between fasting plasma leptin, heart rate variability measures and blood pressure of 68 men and 62 women Men
HR (bpm) SBP (mm Hg) DBP (mm Hg) HF (nU) LF (nU) HF/LF
Fig. 2. Plasma leptin with time following intravenous glucose. Data displayed as mean (se).
pressure. In women, leptin correlates positively with LF power and negatively with the HF/LF ratio, suggesting that leptin is associated with increased cardiac sympathetic and decreased cardiac vagal tone (these relationships remain highly significant allowing for any of the measures of adiposity). Despite this
Women
r
p
r
p
0.10 0.19 0.24 0.06 0.02 0.01
0.419 0.132 0.056 0.617 0.865 0.919
0.27 0.09 0.17 − 0.15 0.42 − 0.33
0.032 0.507 0.200 0.243 0.001 0.008
relationship, there was no association between leptin and blood pressure. While the association between heart rate variability and leptin was not seen in men, a relationship between heart rate and insulin sensitivity was only seen in men [6]. In a stepwise multiple regression analysis in the men, with HF/LF ratio as the outcome variable, only Si (beta = 0.438 and p < 0.001) and supra-iliac skinfold thickness (beta = 0.33 and p = 0.013) remained significant among all metabolic variables and measures of body anthropometry. Only leptin remained significant (beta = − 0.33 and p = 0.008) in the corresponding regression in women. The relationships described above were not confounded by levels of physical activity, alcohol consumption, smoking, stage of the menstrual cycle or hormonal contraceptive use. 4. Discussion
Fig. 3. Relationship between fasting leptin and resting heart rate for 68 men (r = 0.10 and p = 0.419) and 62 women (r = 0.27 and p = 0.032).
Numerous studies have reported gender differences in the physiology of leptin, but it remains unclear whether the dichotomy reflects differences in total body fat, or in body fat distribution [14–16]. Previous work has also indicated an association between leptin and the SNS [17]. Our data confirm this, but find the association to be closer in women than in men. Although there was no gender difference in BMI, women were considerably more adipose than men and their circulating leptin concentrations were four fold higher. BMI is a poor index of fat mass/distribution, and supra-iliac skinfold thickness – a proxy for % body fat – was the strongest predictor of fasting leptin in both sexes. Fasting insulin was a stronger predictor of fasting leptin than insulin sensitivity measured by IVGTT, and in women the waist–hip ratio, an index of central fat distribution, was an additional predictor of fasting leptin. Although fasting insulin is positively associated with fasting leptin, it is not clear which is causal, or indeed whether some cocorrelate is responsible. The rise in leptin following an intravenous bolus of glucose might suggest that insulin drives leptin secretion, a view supported by the rise in circulating leptin reported within five days of starting insulin therapy in patients with newly diagnosed type 1 diabetes [18]. Hyperinsulinemic euglycemic clamp studies, however, show that it is the rate of glucose uptake into adipose tissue rather than the prevailing insulin levels that determine leptin secretion [19]. In vitro experiments support this view [20], though plasma glucose was not an independent predictor of fasting leptin in the present study.
D.E. Flanagan et al. / Regulatory Peptides 140 (2007) 37–42
Resting heart rate is a highly variable measure. Although we carried out this study under standardized conditions, using a relatively long resting ECG (15 min), it was impossible to control completely for anxiety. Although we found strong relationships between heart rate, measures of heart rate variability, leptin and insulin resistance, they are still likely to be underestimates because of measurement ‘noise’. The power spectral analysis of heart rate variability suggests that the relationship between heart rate and metabolic variables is mediated through alterations in cardiac autonomic tone. Although cardiac automaticity is intrinsic, the resting pulse is largely under the control of the autonomic nervous system. Power spectral analysis enables quantitative evaluation of beatto-beat cardiovascular control, and distinguishes the two major components of the spectrum of heart rate variability, namely the high and low frequency components [21]. The high frequency component is centered at the respiratory frequency and is a measure of respiratory sinus arrhythmia. Since respiratory sinus arrhythmia is mediated almost entirely by the vagus nerve, HF power is taken as an index of cardiac vagal control [13]. Low frequency power is thought to be a resonance phenomenon due to delay in the sympathetic feedback loops of the baroreflex system and contains both sympathetic and parasympathetic contributions. Hence, the ratio of HF to LF is considered to represent the balance between vagal and sympathetic contributions to heart rate variability [13,22]. We have previously published evidence of marked gender differences in the relationship between Si and heart rate. Men, but not women, showed a strong relationship between resting heart rate and Si — an increasing heart rate being associated with increasing insulin resistance, independent of BMI. Increasing insulin resistance was also associated with increased sympathetic and decreased parasympathetic cardiac tone [6]. The present data suggest that fasting leptin is more closely related to resting heart rate in young women than young men. Fasting leptin was also correlated with the low frequency component of heart rate variability and the high to low frequency ratio, suggesting that increased circulating leptin is associated with increased cardiac sympathetic and decreased parasympathetic tone. The techniques available for assessing human autonomic activity are often technically difficult or imprecise. We have chosen to measure cardiac autonomic activity because it is non-invasive and technically easy to perform. Prospective studies have shown that measures of heart rate variability are predictive of cardiovascular mortality and morbidity [23]. Previous work suggests that the SNS is an important determinant of leptin secretion. It is known to be important in mobilising fatty acids from adipose tissue [24], and is an important regulator of leptin secretion from adipose tissue. Cold exposure or fasting, both of which raise SNS activity in fat, leads to decreased leptin production [25]. Conversely, sympathetic blockade may lead to increased leptin production [26]. The work by Haynes et al. [3,27] would support the view that leptin is acting to increase sympathetic nervous system activity, this may reflect differing autonomic effects in different tissues. Our results would support the latter view. SNS activity is
41
reduced in rodents with leptin deficiency or leptin receptor defects, contributing to decreased energy expenditure [27]. Obesity is associated with increased measures of SNS activity in humans, and animal work has shown that pharmacological doses of leptin increase SNS activity and blood pressure [28]. One previous study has examined the relationship between heart rate variability and plasma leptin in an older population of men [15]. In this study, mean age of the subjects was 42.5 years and, as would be expected, the subjects were more insulin resistant with higher fasting insulin and leptin concentrations. Interestingly, both HOMA measures of insulin resistance and leptin were predictive of cardiac autonomic activity. This may reflect changing relationships between leptin and the insulin resistance syndrome with time, and illustrates the difficulty in separating the complex relationships between the different systems. We could find no relationship between leptin and blood pressure in our young population. Our observational data would support the hypothesis that leptin may be contributing to the gender differences in the relationship between adiposity and cardiovascular risk — a view that is supported by a number of other groups [29,30]. In summary, we have shown major gender differences in the relationships between leptin, cardiac autonomic activity and insulin/insulin resistance. Importantly, the study was performed in a population of young adults, before features of the insulin resistance syndrome had become established. In women, but not in men, increased leptin is associated with a shift in cardiac autonomic activity towards increased sympathetic tone and greater cardiometabolic risk. The findings may be related to the intrinsically higher insulin resistance recently reported in women. Acknowledgements The authors would like to thank Sister M. Logan, Ms M. Rourke, Ms L. Raggett for assistance with the fieldwork. This study was generously supported by the Medical Research Council, Diabetes UK, the Wessex Medical Trust and the Women's and Children's Hospital Foundation. VMM was supported by the PHRDC, National Health and Medical Research Council of Australia. IFG is supported by the Heart Disease and Diabetes Research Trust. References [1] Zhang Y, Proenca R, Maffei M, Barone M, Leopold L, Friedman JM. Positional cloning of the mouse obese gene and its human homologue. Nature 1994;372:425–32. [2] Leyva F, Godsland IF, Ghatei M, Proudler AJ, Aldis S, Walton C, Bloom S, Stevenson JC. Hyperleptinemia as a component of a metabolic syndrome of cardiovascular risk. Arterioscler Thromb Vasc Biol 1998;18:928–33. [3] Haynes WG, Morgan DA, Walsh SA, Mark AL, Sivitz WI. Receptormediated regional sympathetic nerve activation by leptin. J Clin Invest 1997;100:270–8. [4] Facchini FS, Stoohs RA, Reaven GM. Enhanced sympathetic nervous system activity. The linchpin between insulin resistance, hyperinsulinemia, and heart rate. Am J Hypertens 1996;9:1013–7. [5] Flanagan DE, Moore VM, Godsland IF, Cockington RA, Robinson JS, Phillips DI. Fetal growth and the physiological control of glucose tolerance in adults: a minimal model analysis. Am J Physiol 2000;278:E700–6.
42
D.E. Flanagan et al. / Regulatory Peptides 140 (2007) 37–42
[6] Flanagan DE, Vaile JC, Petley GW, Moore VM, Godsland IF, Cockington RA, Robinson JS, Phillips DIW. The autonomic control of heart rate and insulin resistance in young adults. J Clin Endocrinol Metab 1999;84:1263–7. [7] Bergman RN, Ider Y, Bowden C, Cobelli C. Quantitative estimation of insulin sensitivity. Am J Physiol 1979;236:E667–77. [8] Walton C, Godsland IF, Proudler AJ, Felton C, Wynn V. Evaluation of four mathematical models of glucose and insulin dynamics with analysis of effects of age and obesity. Am J Physiol 1992;262:E755–62. [9] Alpha B, Cox L, Crowther N, Clark PM, Hales CN. Sensitive amplified immunoenzymometric assays (IEMA) for human insulin and intact proinsulin. Eur J Clin Chem Clin Biochem 1992;30:27–32. [10] Burg JP. A new analysis technique for time series data. Erschede Netherlands: NATO Advanced Study Institute on signal processing with emphasis on underwater acoustics; 1968. [11] Akaike H. Statistical predictor identification. Ann Inst Stat Math 1970;22:203–17. [12] Zetterberg LH. Estimation of parameters for a linear difference equation with application to EEG analysis. Math Biosci 1969;5:227–75. [13] Malliani A, Pagani M, Lombardi F, Cerutti S. Cardiovascular neural regulation explored in the frequency domain. Circulation 1991;84:482–92. [14] Ostlund Jr RE, Yang JW, Klein S, Gingerich R. Relation between plasma leptin concentration and body fat, gender, diet, age, and metabolic covariates. J Clin Endocrinol Metab 1996;81:3909–13. [15] Paolisso G, Rizzo MR, Mone CM, Tagliamonte MR, Gambardella A, Riondino M, Carella C, Varricchio M, D'Onofrio F. Plasma sex hormones are significantly associated with plasma leptin concentration in healthy subjects. Clin Endocrinol 1998;48:291–7. [16] Liuzzi A, Savia G, Tagliaferri M, Lucantoni R, Berselli ME, Petroni ML, De Medici C, Viberti GC. Serum leptin concentration in moderate and severe obesity: relationship with clinical, anthropometric and metabolic factors. Int J Obes Relat Metab Disord 1999;23:1066–73. [17] Rayner DV, Trayhurn P. Regulation of leptin production: sympathetic nervous system interactions. J Mol Med 2001;79:8–20. [18] Soliman AT, Omar M, Assem HM, Ibrahim SN, Rizk MM, El Matary W, El Alaily RK. Serum leptin concentrations in children with type 1 diabetes mellitus: relationship to body mass index insulin dose and glycemic control. Metabolism 2002;51:292–6. [19] Wellhoener P, Fruehwald-Schultes B, Kern W, Dantz D, Kerner W, Born J, Fehm HL, Peters A. Glucose metabolism rather than insulin is a main
[20] [21]
[22]
[23]
[24]
[25]
[26]
[27] [28]
[29]
[30]
determinant of leptin secretion in humans. J Clin Endocrinol Metab 2000;85:1267–71. Levy JR, Stevens W. The effects of insulin, glucose, and pyruvate on the kinetics of leptin secretion. Endocrinology 2001;142:3558–62. Anonymous. Heart rate variability. Standards of measurement, physiological interpretation, and clinical use. Task Force of the European Society of Cardiology and the North American Society of Pacing and Electrophysiology. Eur Heart J 1996;17:354–81. Akselrod S, Gordon D, Ubel FA, Shannon DC, Barger AC, Cohen RJ. Power spectrum analysis of heart rate fluctuations: a quantitative probe of beat to beat cardiovascular control. Science 1981;213:220–2. La Rovere MT, Bigger Jr JT, Marcus FI, Mortara A, Schwartz PJ, ATRAMI (Autonomic Tone and Reflexes After Myocardial Infarction) Investigators. Baroreflex sensitivity and heart-rate variability in prediction of total cardiac mortality after myocardial infarction. Lancet 1998;351:478–84. Millet L, Barbe P, Lafontan M, Berlan M, Galitzky J. Catecholamine effects on lipolysis and blood flow in human abdominal and femoral adipose tissue. J Appl Physiol 1998;85:181–8. Trayhurn P, Duncan JS, Rayner DV. Acute cold-induced suppression of ob (obese) gene expression in white adipose tissue of mice: mediation by the sympathetic system. Biochem J 1995;311:729–33. Evans BA, Agar L, Summers RJ. The role of the sympathetic nervous system in the regulation of leptin synthesis in C57BL/6 mice. FEBS Lett 1999;444:149–54. Haynes WG. Interaction between leptin and sympathetic nervous system in hypertension. Curr Hyperten Rep 2000;2:311–8. Satoh N, Ogawa Y, Katsuura G, Numata Y, Tsuji T, Hayase M, Ebihara K, Masuzaki H, Hosoda K, Yoshimasa Y, Nakao K. Sympathetic activation of leptin via the ventromedial hypothalamus: leptin-induced increase in catecholamine secretion. Diabetes 1999;48:1787–93. Lyoussi B, Ragala MA, Mguil M, Chraibi A, Israili ZH. Gender specific leptinaemia and its relationship with some components of the metabolic syndrome in Moroccans. Clin Exp Hypertens 2005;27:377–94. Couillard C, Mauriege P, Prud’homme D, Nadeau A, Tremblay A, Bouchard C, Despres JP. Plasma leptin concentrations: gender differences and associations with the metabolic risk factors for cardiovascular disease. Diabetologia 1997;40:1178–84.