Androgens and selected cardiovascular risk factors in

Androgens and selected cardiovascular risk factors in aging men

Hamid Reza Nakhai-Pour

Androgens and selected cardiovascular risk factors in aging men Hamid Reza Nakhai-Pour

Thesis, Utrecht University, Faculty of Medicine, with a summary in Dutch. Proefschrift, Universiteit Utrecht, met een samenvatting in het Nederlands.

ISBN

978-90-393-4552-8

Author


Lay out


Print

Gildeprint Drukkerijen B.V., Enschede, the Netherlands

Utrecht 2007

© HR. Nakhai-Pour, 2007 All rights reserved. No part of this thesis may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopy, recording, any information storage and retrieval system or otherwise, without prior written permission of the author. The copyrights on articles that have been published or accepted for publication have been transferred to the respective journals.

Androgens and selected cardiovascular risk factors in aging men

Androgenen en cardiovasculaire risicofactoren bij oudere mannen (met een samenvatting in het Nederlands)

Proefschrift

ter verkrijging van de graad van doctor aan de Universiteit Utrecht op gezag van de rector magnificus, prof.dr. W.H. Gispen, ingevolge het besluit van het college voor promoties in het openbaar te verdedigen op donderdag 12 juli 2007 des morgens te 10.30 uur

door

Hamid Reza Nakhai-Pour geboren op 23 juni 1968 te Tehran, Iran

Promotor:

Prof. dr. D.E. Grobbee

Co-promotor:

Dr. ir.Y.T. van der Schouw Dr. Harald JJ Verhaar

Financial support by the Netherlands Heart Foundation for the publication of this thesis gratefully acknowledged. Additional financial support was also provided by the Netherlands Geriatrics Research Foundation, Iranian Ministry of Health and Medical Education, UMC Utrecht, Organon N.V., GlaxoSmithKline B.V. and Servier Nederland Farma B.V.

To Gisso and Kiana; And in loving memory of my father

Contents_______________________________________________________________________

Chapter 1

General Introduction ………………………....…………………………

9

Chapter 2

Circulating androgens and cardiovascular risk factors………....

13

2.1 Androgens levels and circulating homocysteine…………………

15

2.2 Androgens levels and chronic low-grade inflammation ………………..

29

Large artery stiffness and cardiovascular risk factors …………………

41

3.1 Large artery stiffness and endogenous androgens levels …………

43

3.2 Large artery stiffness and chronic low-grade inflammation ……….

55

3.3 Large artery stiffness and thickness and circulating homocysteine levels

69

Testosterone supplementation and cardiovascular risk factors

83

Chapter 3

Chapter 4

4.1 Trial design and baseline characteristics of a double-blind randomized placebo-controlled trial on the effects of testosterone supplementation in elderly men with moderate low testosterone levels..

85

4.2 Effect of oral testosterone supplementation on blood pressure and aortic artery stiffness; 6 months randomized controlled trial …………

105

4.3 Effect of oral testosterone supplementation on chronic low-grade inflammation in older men; 26 weeks randomized controlled trial ……

Chapter 5

General discussion

………………………………………….

117

127

Summary ………………………………………………………………..

133

Summary in Dutch / Nederlandse samenvatting ……………………

137

Acknowledgement ……………………………………………………..

141

Curriculum Vitae ………………………………………………………..

145

List of publications ……………………………… ……………………..

149

Chapter 1. General introduction

9

Chapter 1 General background ______________________________________________________________ Cardiovascular diseases (CVDs) are the leading cause of death and major determinants of chronic disability among elderly in developed countries 1. Several factors accelerate the rate of progression of atherosclerosis and consequently the incidence of cardiovascular events, such as increasing age, hypercholesterolemia, hypertension, cigarette smoking, diabetes mellitus, obesity, and male gender 2. As the understanding of the processes leading to clinical cardiovascular events has improved, new risk factors have emerged. It is now widely appreciated that as many as half of persons experiencing a first myocardial infarction or cardiac sudden death do not have a high level of risk, as estimated by the standard risk factors (smoking, hypertension, obesity, lack of physical activity, LDL and HDL cholesterol) that are routinely used to estimate likelihood of cardiovascular disease. Circulating sex hormones, homocysteine and evidence of chronic low-grade inflammation have been proposed as novel cardiovascular risk factors that could improve the prediction of cardiovascular events in these cases. In the last decade more attention has been drawn to the potential importance of androgens in etiology, prevention and treatment of male cardiovascular disease. The prevalence of men with low testosterone levels increase dramatically with age. While 12% of men aged less than 50 years have low serum total testosterone levels, this may amount to 50% in age 80 3. It has been suggested that with aging the effects of circulating androgens levels are of increasing importance for the risk of cardiovascular disease 4. Although the exact mechanism by which androgens levels affect the risk is not well understood, but several hypotheses have been proposed. There are observations to suggest that androgens may be involved in the pathogenesis of atherosclerosis, either as independent risk factors or mediated via traditional or newer risk factors, such as diabetes, obesity, hypertension, hypercholesterolemia5, homocysteine 6,7, chronic low-grade inflammation 8,9. Results from studies are, however, conflicting. Definitive conclusions about the relation between endogenous androgens and the aforementioned risk factors cannot be drawn because of selection in study populations, a small sample size, a lack of control for important cardiovascular risk factors and a short duration of androgen supplementation in several of the previously conducted studies. The aim of this thesis was to obtain more insight in the relations between androgens and selected cardiovascular risk factors in men. For this purpose we conducted two studies, a crosssectional study in a population-based sample of 400 middle-aged and elderly men and a randomized controlled trial in 237 elderly men. In the cross-sectional study we investigated the associations between endogenous androgens and circulating homocysteine, C-reactive protein and large artery stiffening as an approximation for cardiovascular end-organ damage. Furthermore we studied the associations between large artery stiffness and thickness and homocysteine as well as large artery stiffness and C-reactive protein levels. The placebo-controlled randomized trial in 237 elderly men with moderate to low testosterone levels was conducted to assess the effect of oral testosterone supplementation on chronic low-grade inflammation, blood pressure measurements (systolic blood

10

General discussion pressure, diastolic blood pressure, and pulse pressure) and large artery stiffness as an approximation of cardiovascular disease risk. Outline of this thesis_ ___________________ ________________________________________ Chapter 2 presents our findings on the association of endogenous androgens with selected cardiovascular risk factors in 400 middle-aged and elderly men from the HAMLET cohort. The association of circulating androgens levels with plasma homocysteine levels is described in chapter 2.1. Furthermore, circulating endogenous estradiol levels were related to C-reactive protein levels in this cohort (chapter 2.2). In chapter 3 the associations between large artery stiffness, endogenous androgens and selected cardiovascular risk factors such as homocysteine and C-reactive protein were accessed. Chapter 4 addresses the intervention study to investigate the effects of testosterone supplementation on functional mobility, quality of life, body composition, cognitive function, vascular function and risk factors, and bone mineral density in elderly men with moderate to low testosterone levels. The design and baseline characteristic of the double-blind, randomized placebo-controlled trial are described in chapter 4.1. The effect of testosterone supplementation on blood pressure and large artery stiffness is shown in chapter 4.2 and the effects on C-reactive protein levels are described in chapter 4.3. In chapter 5 the results described in this thesis are briefly summarized. Some methodological problems are discusses. Finally, an outline of possible future research directions in testosterone supplementation research is presented. References_ ____________________________

______________________________________

1.

Leest LATM, Koek HL, Trip MJCA van, et al. Hart-en vaatziekten in Nederland 2005; cijfers over risicofactoren, ziekte, behandeling en sterfte. Nederlandse Hartstichting 2005.

2.

Criqui M. Epidemiology of atherosclerosis: an updated overview. Am J Cardiol 1986; 57:18C-23C.

3.

Harman SM, Metter EJ, Tobin JD, Pearson J, Blackman MR. Longitudinal effects of aging on serum total and free testosterone levels in healthy men. J Clin Endocrinol Metab 2001; 86:724-731.

4.

Abbott RD, Curb JD, Rodriguez BL et al. Age-related changes in risk factor effects on the incidence of coronary heart disease. Annals of Epidemiology 2002; 12:173-181.

5.

Muller M, van der Schouw YT, Thijssen JHH, Grobbee DE. Endogenous sex hormones and cardiovascular disease in men. J Clin Endocrinol Metab 2003; 88:5076-5086.

6.

Giltay EJ, Hoogeveen EK, Elbers JMH, Gooren LJG, Asscheman H, Stehouwer CDA. Effects of sex steroids on plasma total homocysteine levels: a study in transsexual males and females. J Clin Endocrinol Metab 1998; 83:550553.

11

Chapter 1 7.

Ebenbichler CF, Kaser S, Bodner J, Gander R, Lechleitner M, Herold M, Patsch JR. Hyperhomocysteinemia in bodybuilders taking anabolic steroids. Eur J Intern Med 2001; 12:43-47.

8.

Laaksonen DE, Niskanen L, Punnonen K, et al. Sex hormones, inflammation and the metabolic syndrome: a population-based study. Eur J Endocrinol. 2003; 149:601-608.

9.

Kovacs A, Henriksson P, Hamsten A, Wallen H, Bjorkegren J, Tornvall P. Hormonal regulation of circulating Creactive protein in men. Clin Chem 2005; 51:911-913.

12

Chapter 2. Circulating androgens and selected cardiovascular risk factors

13

14

Chapter 2.1. Androgens levels and circulating homocysteine.

Manuscript based on this chapter: Nakhai Pour HR, Grobbee DE, Muller M, Emmelot-Vonk MH and van der Schouw YT. Serum sex hormone and plasma homocysteine levels in middle-aged and elderly men. European Journal of Endocrinology. 2006; 155:887-893.

15

Chapter 2 Summary________________________________________________________________________ Objective: To investigate whether circulating levels of testosterone (total, bioavailable), estradiol (total, bioavailable), and dehydroepiandrosterone sulphate (DHEAS) are associated with fasting plasma homocysteine (tHcy) levels in middle-aged and elderly men. Subjects and methods: A population-based sample of 400 independently living men between 40 and 80 years of age in a cross-sectional study has been studied. Total testosterone, sex hormone binding globulin (SHBG) and total estradiol were measured by radioimmunoassay methods and bioavailable testosterone and estradiol were calculated. DHEAS was measured using an immunometric technique. Fasting homocysteine was measured by Fluorescence Polarization Immunoassay. Anthropometric characteristics were also measured and two standardized questionnaires completed, including lifestyle factors and diet. Linear regression analysis adjusted for age, BMI, creatinine clearance and mean visceral fat was used to assess the association of endogenous sex hormones and fasting plasma homocysteine levels. Results: After adjustment for age, BMI, creatinine clearance, and mean visceral fat no statistically significant association was observed between testosterone (total, bioavailable), DHEAS, and estradiol (total, bioavailable)levels with natural log tHcy (β = -2×10-3; 95% confidence interval (CI) -9×10-3; 5×10-3), (β = -4×10-3; 95% CI -18×10-3; 9×10-3), (β =3×10-3; 95% CI -6×10-3; 12×10-3), (β = -9.3×10-5; 95% CI 1×10-3; 1×10-3) and (β = 0.00; 95% CI -3×10-3; 2×10-3) respectively. Additional adjustment for smoking, alcohol intake, daily physical activity, diabetes mellitus, and hypertension did not change these findings. Conclusion: The results of our study do not support a direct role for circulating sex hormone levels in the regulation of fasting plasma tHcy concentrations in middle-aged and elderly men.

16

Androgens and homocysteine Introduction______________________________________________________________________ Endogenous androgens are increasingly being investigated in relation to cardiovascular risk factors and as potential predictors of health problems affecting men 1.Testosterone and estradiol levels (total, free and non-SHBG-bound) significantly decline with ageing, whereas SHBG levels increase, making the effective decrease in free and bioavailable testosterone and estradiol even more pronounced 2. Other important androgenic hormones demonstrating age-related declines are the circulating levels of dehydroepiandrosterone (DHEA) and its sulphate (DHEA-S) 3,4. Many studies have demonstrated that lower androgen levels are associated with an increased risk of cardiovascular disease 5,6. Others however found no relationship 7-10. Increased levels of plasma total homocysteine (tHcy) are also common in elderly people 11, 12. Plasma tHcy has shown to be a strong and independent risk factor for cardiovascular disease 13, 14, although the etiological meaning of this finding is less clear, since recent studies indicate that lowering tHcy levels does not decrease cardiovascular disease risk 15,16 A few prior studies have reported data on the probable association of various endogenous sex hormones and tHcy levels. It has been shown that men have higher tHcy concentrations than women-typically about 10 to 15% higher levels but this difference is attenuated after menopause 17. One study found that tHcy levels are typically lower during pregnancy, a state characterized by high levels of endogenous estrogens 18. Moreover, plasma tHcy concentrations were high in patients with polycystic ovary syndrome, a condition characterized by androgen excess 19. In addition, a study in transsexual males and females showed that tHcy levels decrease after estrogens and anti-androgen administration to male (transsexual) subjects 20. Androgen supplementation has been found to induce hyperhomocysteinemia in men taking anabolic androgenic steroids, and also in men with Klinefelter’s syndrome when given supraphysiologic doses of testosterone 21,22. These evidences could be mediated by possible association between endogenous sex hormone and tHcy levels. The aim of present study was to determine the relation of endogenous testosterone (total, bioavailable), total estradiol, DHEA-S levels and fasting plasma tHcy levels in middle-aged and elderly men. Subjects and methods_____________________________________________________________ We conducted a cross-sectional study in 400 men aged 40-80 years apparently healthy at time of enrollment. Details of the study design, recruitment, and procedures have been published 23,24. The study was approved by the institutional Review Board of the University Medical Center Utrecht and written informed consent was obtained from all participants.

17

Chapter 2 Homocysteine measurement Fasting plasma concentration of tHcy was determined as free plus protein-bound tHcy by Fluorescence Polarization Immunoassay on an IMx Analyzer (Abbott Diagnostics, Hoofddorp, The Netherlands). For variation reduction in plasma tHcy measurement, sampling was done in standardized sitting down position. The samples were coagulated with EDTA and centrifuged after collection 25. The inter-assay variation is 3.9; 4.2 and 3.9% at 9.9; 18.1 and 41.0 µ mol/l respectively (n=73) and the lower limit of detection was 0.5µ mol/l. Sex hormone measurements A venapuncture was performed between 08:00-10:00h am after overnight fasting, and a blood sample was obtained. Platelet free serum and plasma was obtained by centrifugation and immediately stored at -20°C. Circulating androgens levels were measured in serum. Total testosterone (TT) was measured after diethyl ether extraction using an in-house competitive radioimmunoassay employing a polyclonal anti-testosterone-antibody (Dr Pratt AZG 3290). The lower limit of detection was 0.24 nmol/l and inter-assay variation was 6.0; 5.4 and 8.6% at 2.1; 5.6 and 23 nmol/l respectively (n=85). Bioavailable testosterone (free plus albumin bound) was calculated from measures of total testosterone and immunoassayed SHBG concentrations and albumin using the method described by Vermeulen et. al. 26. Estradiol (E2) was measured after diethyl ether extraction and Sephadex chromatography using an in-house competitive radioimmunoassay employing a polyclonal anti-estradiol-antibody (Dr F de Jong, Erasmus MC, Rotterdam NL). The lower limit of detection was 20 pmol/l and inter-assay variation was 10.0 at 81 pmol/l and 3.1 % at 660 pmol/l (n=24, resp.17). Bioavailable E2 was calculated using the method described by Södergard et al. 27. This method is based on the knowledge of the total concentration of all steroids competing for the same binding site on SHBG, the concentration of albumin (using a fixed concentration of 40 gr/l), the binding capacity of SHBG, and the association constant of E2 to the binding proteins. Dehydroepiandrosterone-sulfate (DHEA-S) was measured using an immunometric technique on an advantage chemiluminescense system (Nichols Institute Diagnostics, San Juan Capistrano, USA). The lower limit of detection was 0.1µ mol/l and inter-assay variation was 5.2; 5.6 and 4.2% at 1.0; 4.9 and 14.2 µ mol/l respectively (n=19). Sex hormone binding globulin (SHBG) was measured using an immunometric technique on an IMMULITE Analyser (Diagnostic Products Corporation, Los Angeles, USA). The lower limit of detection was 5 nmol/l and inter-assay variation was 6.1% at 11.6 nmol/l, 4.9% at 36 nmol/l and 6.9% at 93 nmol/l (n=30). Other measurements A trained physician obtained information on prevalent diseases, medication use, and life-style factors. Smoking was estimated from self-report and categorized in current, former, and never smokers. Alcohol consumption was assessed by a validated Food Frequency Questionnaire (FFQ), designed

18

Androgens and homocysteine to estimate regular intake of 178 food items in the year prior to enrolment 28. Height and weight were measured in standing position without shoes. Waist and hip circumference were measured using a standard household centimeter and rounded to the nearest centimeter. Waist-to-hip ratio was calculated by dividing waist circumference by hip circumference. Blood pressure (BP) was measured twice at the right brachial artery with a semi-automated device (Dinamap, GE HealthCare, Tampa, FL). The average of the two measurements of systolic and diastolic blood pressure was used for analysis and further calculation. Body mass index (BMI) was calculated at the weight in kilograms divided by the square of the height in meters (kg/m2). Visceral and subcutaneous fat were measured using ultra-sound measurement 29. Ultrasonography was performed with an HDI 3000 (Philips Medical Systems, Eindhoven, the Netherlands) using a C 4-2 transducer. For all images, the transducer was placed on a straight line drawn between the left and right midpoints between the lower rib and iliac crest. Measurements were made at the end of quiet expiration, applying minimal pressure without displacement of intraabdominal contents as observed by ultrasound image. Visceral fat comprises the distance between peritoneum and lumbar spine. Subcutaneous fat comprises the distance between the skin and the linea alba. For the measurement of visceral fat mass, we measured distances from three different angles (medial, left and right). For the subcutaneous fat mass the distance from the medial angle was measured in threefold. The mean of the three different measurements was calculated and used in our analyses 30. Physical activity was assessed using a questionnaire validated in an elderly population, and was categorized in quartiles 31. Fasting glucose levels were assessed using a GlucoTouch reflectometer (LifeScan, Inc., Benelux, Beese, Belgium), a reagent –strip glucose oxidase methods. Serum creatinine was measured by the Jaffe rate method (Synchron LX® i 725, Beckman Coulter, Inc). Creatinine clearance was calculated from serum creatinine, age, and body weight according to the Cockcroft-Gault formula (140-age (yr) × Weight (kg) × constant / serum creatinine (μmol/l)) 32. An automatic enzymatic procedure was used to determine serum total cholesterol (Synchron LX Systems; Beckman Coulter, Inc). High-density lipoprotein (HDL) cholesterol and triglycerides were measured similarly. Diabetes mellitus was defined as treatment with insulin or oral hypoglycaemic agents or fasting plasma venous glucose > 6.9 mmol/l. Hypertension was defined as systolic blood pressure ≥160 mmHg and/or diastolic blood pressure ≥ 95 mmHg and/or use of anti-hypertensive medication. Hyperlipidemia was defined as serum total cholesterol > 6.5 mmol/l and/or treatment with lipidlowering medication. The subjects’ customary alcohol intake, caffeine use, daily intake of vitamin B groups, protein, methionine and folate, were estimated from a validated food frequency questionnaire. Presence of cardiovascular disease was defined as coronary heart disease, peripheral artery disease or stroke.

19

Chapter 2 Data analysis We excluded the subjects with CVD (n= 58) because it has been shown that patients with prevalent cardiovascular disease have higher homocysteine levels and lower sex hormone levels than patients without. Therefore, inclusion of subjects with prevalent disease may lead to spurious correlation. We examined the normality assumptions for the variables with the Kolmogorov-Smirnov test. Because the tHcy and triglyceride concentrations were skewed, we used log-transformed values. For alcohol consumption as data suggested the presence of a J-shaped relation, therefore alcohol intake was divided into four levels (0-3, 4-10, 11-21 and 22-higher unit per week) and the first group was chosen as the reference group in the analysis. The alcohol intake levels and smoking were put into the model as dummy variables. The association between circulating serum sex hormone levels and natural log plasma tHcy was examined using linear regression coefficients (β) and corresponding 95% confidence intervals (CI). To determine whether and to what extent the observed associations of sex hormones levels with might be explained by intermediate factors, multivariate regression analysis were used with adjustment for age, BMI, creatinine clearance and mean visceral fat, as based on literature indicating there to be major determinants of circulating sex hormones and plasma tHcy (Model A). Further adjustment has been done for smoking, alcohol intake and daily physical activity (Model B) and hypertension and diabetes mellitus (model C), respectively. Because of the large effect of age and obesity (waist-to-hip ratio ≥ 0.98) on plasma tHcy and sex hormone levels, effect modification was tested by adding the interaction terms to the multivariate model. Stratified analyses were presented when the interaction term was statistically significant (P-value < 0.05). The data were analyzed using the SPSS statistical package for windows version 11.0 (SPSS Inc., Chicago, IL, USA). Results__________________________________________________________________________ The mean age of the population was 60 years, and the median of fasting tHcy was 12.10 µ mol/l (9.40-17.50) (Table1). 21.3% and 0.8% of subjects showed mild (15-16 µ mol/l) and intermediate (16-30 µ mol/l) hyperhomocysteinemia, respectively. Subjects with the highest bioavailable testosterone level were younger, less likely to have cardiovascular disease, hypertension, diabetes mellitus, and more likely to be current smoker. General characteristics of study subjects across the quartiles of bioavailable testosterone levels are shown in table 2.

20

Androgens and homocysteine

Table 1. Characteristics of the study population (n=342).

Characteristic Age (Years) Weight (Kg) BMI (Kg/m2) Waist-to-hip ratio Systolic blood pressure (mmHg) Diastolic blood pressure (mmHg) Visceral fat (cm) Subcutaneous fat (cm) Visceral and subcutaneous fat (cm) Serum creatinine (µmol/l) Creatinine clearance (ml/min) Total cholesterol (mmol/l) Glucose (mmol/l) Physical activity (Voorrips score) Hormone levels Total Testosterone (nmol/l) Bioavailable Testosterone (nmol/l) Total Estradiol (pmol/l) Bioavailable Estradiol (pmol/l) SHBG (nmol/l) DHEAS (µ mol/l) Homocysteine(µ mol/l)

Mean±S.D 60.2±11.3 83.22±12.20 26.27±3.47 0.97±0.06 143.4±22.1 81.5±10.3 7.54±2.23 2.64±0.85 10.09±2.45 90.0±14.2 1.07±0.29 5.89±1.03 5.96±1.45 18.06±7.46 18.53±5.33 8.14±2.25 91.25±22.84 42.19±11.44 40.58±14.46 6.67±3.28 13.20±4.01 Median (percentiles 10-90)

Triglyceride (mmol/l) Alcohol use(units/week) Daily caffeine use(mg) Homocysteine (µ mol/l)

1.20(0.70-2.72) 10(0.0-31.8) 399.2(166.6-641.7) 12.10(9.40-17.50) N (%)

Hypertension Diabetes mellitus Alcohol use(units/week) Current smoking

79(23.0) 77(23.0) 86(25.5) 79(23)

21

22 Alcohol use (unit/week) 0-3 4-10 11-21 22≤ Current smoking

Hypertension Diabetes mellitus

Triglyceride (mmol/l) Alcohol use (unit/week) Daily caffeine use (mg) Homocysteine (µ mol/l)

Bioavailable testosterone (nmol/l) Age (years) SHBG (nmol/l) Systolic Blood Pressure (mmHg) Diastolic Blood Pressure (mmHg) Total folate intake (mµ g) BMI (Kg/m2) Waist-to-hip ratio Visceral fat (cm) Subcutaneous fat (cm) Visceral and subcutaneous fat (cm) Serum creatinine (μmol/l) Creatinine clearance (ml/min) Total cholesterol (mmol/l) Physical activity(Voorrips score)

26 (33) 15 (19) 25 (32) 12 (15) 15 (19)

Quartile1 (n=79) Mean ±SD 5.56±0.97 66.58±8.79 30.21±10.04 150.32±25.05 83.28±9.40 222.67±75.13 26.09±2.93 0.99±0.06 8.06±2.62 2.82±0.98 10.78±2.98 91.1±15.8 0.99±0.30 5.52±1.06 18.4±7.2 Median 1.20 (0.79-2.53) 9.00 (0.00-34.00) 422.00 (110.00-631.00) 12.45 (9.49-18.97) N (%) 28 (36) 12 (15) 20 (24) 24 (29) 17 (21) 21 (25) 18 (22)



Quartile4 (n=91) Mean ±SD 11.20±1.46 51.22±8.86 52.06±15.44 137.03±18.00 80.34±9.64 210.94±53.99 26.56±3.45 0.94±0.06 6.73±1.93 2.48±0.91 9.20±2.18 89.7±12.4 1.15±0.21 5.90±1.09 18.1±7.2 Median 1.20 (0.70-2.29) 11.00 (0.00-30.10) 357.00 (158.00-712.00) 11.85 (8.91-15.49) N (%) 6 (7) 3 (3)

Table2. Characteristics of the study population according to the different quartiles of bioavailable testosterone levels.

0.07 0.03 0.33 0.33 0.07

0.00 0.05

0.00 0.00 0.08 0.00 0.13 0.43 0.54 0.00 0.00 0.05 0.00 0.70 0.00 0.01 0.36

P-value

Chapter 2

Androgens and homocysteine In the unadjusted model, natural log homocysteine decreased significantly with increasing testosterone (total, bioavailable) levels; natural log homocysteine was 8×10-3(95% CI -14×10-3; -2×103) µ mol/l and 17×10-3 (95% CI -28×10-3; -5×10-3) µ mol/l lower for each one nmol/l increase in testosterone (total, bioavailable) levels, respectively. Multiple regression analysis adjusted for age, BMI, creatinine clearance and mean visceral fat showed no statistically significant association between testosterone (total, bioavailable), DHEAS and estradiol (total, bioavailable) levels with natural log tHcy (β = -2×10-3; 95% CI -9×10-3; 5×10-3), (β = -4×10-3; 95% CI -18×10-3; 9×10-3), (β =3×10-3; 95% CI -6×10-3; 12×10-3), (β = -9.3×10-5; 95% CI -1×10-3; 1×10-3) and (β = 0.00; 95% CI 3×10-3; 2×10-3) respectively (table3). Further adjustment for smoking, alcohol intake and daily physical activity diabetes mellitus, hypertension did not materially alter the results (β = -2×10-3; 95% CI -10×10-3; 5×10-3), (β = -4×10-3; 95% CI -19×10-3; 10×10-3), (β =2×10-3; 95% CI -8×10-3; 12×10-3), (β = 0.00; 95% CI -1×10-3; 1×10-3) and (β = 0.00; 95% CI -3×10-3; 2×10-3) respectively. Table 3. Regression coefficients from linear regression models including natural log homocysteine level as dependent variable and endogenous sex hormone levels as independent variable.

Regression coefficient (µmol/l) 95% CI Total testosterone (nmol/l)* Crude -8×10-3 (-14×10-3; -2×10-3) Model A1 -2×10-3 (-9×10-3; 5×10-3) Model B2 -2×10-3 (-10×10-3; 5×10-3) Model C3 -2×10-3 (-10×10-3; 5×10-3) Bioavilable Testosterone (nmol/l) Crude -17×10-3 (-28×10-3; -5×10-3) Model A -4×10-3 (-18×10-3; 9×10-3) Model B -5×10-3 (-19×10-3; 9×10-3) Model C -4×10-3 (-19×10-3;10×10-3) DHEAS (µ mol/l) Crude -10×10-3 (-18×10-3;-1×10-3) Model A 3×10-3 (-6×10-3; 12×10-3) Model B 3×10-3 (-7×10-3; 12×10-3) Model C 2×10-3 (-8×10-3; 12×10-3) Total E2 (pmol/l)* Crude 0.00 (-1×10-3; 1×10-3) Model A -9.3×10-5 (-1×10-3; 1×10-3) Model B 0.00 (-1×10-3; 1×10-3) Model C 0.00 (-1×10-3; 1×10-3) Bioavailable E2(pmol/l) Crude -1×10-3 (-4×10-3; 1×10-3) Model A 0.00 (-3×10-3; 2×10-3) Model B 0.00 (-3×10-3; 2×10-3) Model C 0.00 (-3×10-3; 2×10-3) 1 adjusted for age, BMI, creatinine clearance and mean visceral fat, 2 adjusted for age, BMI, creatinine clearance, mean visceral fat, smoking, alcohol intake, and physical activity. 3 adjusted for age, BMI, creatinine clearance, mean visceral fat, smoking, alcohol intake, physical activity, diabetes mellitus, and hypertension. *additional adjustment for SHBG

23

Chapter 2 Discussion_______________________________________________________________________ In this population-based study, we found no support that circulating testosterone (total, bioavailable), estradiol, or dehydroepiandrosterone sulphate levels and fasting plasma tHcy are independently related in middle-aged and elderly men. Our result in this study is in some part in agreement with results of previously conducted studies, which showed no association between fasting plasma tHcy and androgen supplementation in healthy men. In a crossover study which has been conducted to assess the effects of supraphysiologic does of testosterone, with or without aromatization to estradiol in eugonadal men, it has been shown that fasting tHcy levels were not significantly altered during either testosterone condition, suggesting that testosterone and its aromatization to estradiol does not affect tHcy concentrations 33. A case-control study of patients with idiopathic hypogonadotropic hypogonadism (IHH) treated with gonadotropin and controls, showed that plasma tHcy concentrations were independent of changes in androgen levels 34. Similar to present finding, In a prospective clinical study it has been shown that supraphysiologic doses of testosterone increased plasma tHcy levels in men with Klinefelter’s syndrome, potentially through changes in creatinine, cobalamin and folate 22. However, in a study performed in transsexual males and females subjects, plasma tHcy levels increased after androgen administration to female (trans-sexual) subjects, and substantially decreased in male (trans-sexual) subjects after estrogen and antiandrogen administration, respectively 20. Similarly, in young bodybuilders the intake of anabolic androgenic steroid induced acute Hyperhomocysteinemia 21. The best explanation for the observed gender difference in plasma tHcy concentration is the relation of tHcy concentrations with lean body mass, protein turnover, and creatinine production, all of which tend to be higher in men than women 35. The relation between circulating concentrations of tHcy and serum creatinine, as one of the most important determinants of fasting homocysteine, may largely reflect the effect of renal function on homocysteine concentrations, but tHcy and creatinine could be also be related because of increased homocysteine production during creatinine metabolism. In female to male transsexuals, in addition to plasma creatinine levels, the creatinine excretion rate, lean body mass and body mass index increased. Therefore, sex steroid induced change in plasma tHcy levels could conceivably be explained by their anabolic/ catabolic effects. Exogenous estrogen in elderly men has been suggested to reduce plasma tHcy levels 36. Previous studies have suggested a protective role for endogenous estrogens in cardiovascular system in men 37. Furthermore, estradiol supplementation has been shown to improve endothelial function in both elderly hypogonadal men and healthy men receiving testosterone 38,39. It has been suggested that suppression of endogenous estrogens with an aromatase inhibitor resulted in a decrease in the estradiol level and an impairment of endothelia function, determined by flow mediated dilatation of the brachial artery, without significant changes in lipoproteins or homocysteine 40. It has been suggested that estradiol prevents Hcy-induced endothelial damage in male rats 41. The most important beneficial effects of estradiol on cardiovascular disease are opposite of those harmful

24

Androgens and homocysteine effects induced by Hcy. Furthermore, it has been found in this study that the beneficial effects of estradiol on endothelium were related closely to lower serum tHcy concentrations. DHEAS is an abundant steroid hormone that is a precursor for both estrogens and testosterone, with unknown biological functions. Several clinical and population-based studies suggest that DHEAS play a protective role against atherosclerosis and coronary artery disease in men 42,43. One study found that chronic DHEA supplementation exerts anti-atherogenic effects, although homocysteine concentrations were not modified by the treatment 44. It seems likely that the anti-atherogenic effects of DHEAS are independent of its action on plasma homocysteine concentrations. It seems likely that the anti-atherogenic effects of DHEAS are independent of its action on plasma homocysteine concentrations. The present study has some limitations. Homocysteine is metabolized by remethylation to methionine or by transsulfuration to cysteine. The remethylation pathway regulates fasting levels of homocysteine, whereas, the methionine-load test is a measure the activity of the tHcy catabolic transsulfuration pathway. It has been shown that fasting tHcy levels are only weakly related to the tHcy response to methionine loading 45, therefore we cannot exclude the possibility that endogenous androgens influence the tHcy catabolic transsulfuration pathway. In addition, the interpretability of the results may be restricted by several factors inherent to the cross-sectional design and do not provide direct evidence of cause and effect. Another limitation is that because of within subject biological variation, the single measurement of both sex hormones and several risk factors will reflect long-term averages less precisely than repeated measurements. In summary, we found no significant association of testosterone (total, bioavailable), estradiol (total, bioavailable), or dehydroepiandrosterone sulphate levels with fasting plasma homocysteine levels in middle-aged and elderly men. References_______________________________________________

_____________________

1

Morley JE, Kaiser FE, Sih R, Hajjar R, Perry HM 3rd. Testosterone and frailty. Clin Geriatr Med 1997; 13: 685-695.

2

Feldman HA, Longcope C, Derby CA et al. Age trends in the level of serum testosterone and other hormones in middle-aged Men: longitudinal results from the massachusetts male aging study. J Clin Endocrinol Metab 2002; 87: 589-598.

3

Hinson J, Raven P. DHEA deficiency syndrome: a new term for old age? J Endocrinol 1999; 163:1-5.

4

Simpson ER, Davis SR. Minireview: aromatase and the regulation of estrogen biosynthesis--some new perspectives. Endocrinology 2001; 142:4589-4594.

5

Alexandersen P, Haarbo J, Christiansen C. The relationship of natural androgens to coronary heart disease in males: a review. Atherosclerosis 1996; 125:1-13.

25

Chapter 2 6

Muller M, Van Den Beld AW, Bots ML, Grobbee DE, Lamberts SWJ, van der Schouw YT. Endogenous sex hormones and progression of carotid atherosclerosis in elderly men. Circulation 2004; 109:2074-2079.

7

Barrett-Connor E, Khaw KT. Endogenous sex hormones and cardiovascular disease in men. A prospective population-based study. Circulation 1988; 78: 539-545.

8

Cauley JA, Gutai JP, Kuller LH, Dai WS. Usefulness of sex steroid hormone levels in predicting coronary artery disease in men. Am J Cardiol 1987; 60:771-777.

9

Phillips GB, Yano K, Stemmermann GN. Serum sex hormone levels and myocardial infarction in the Honolulu Heart Program. Pitfalls in prospective studies on sex hormones. J Clin Epidemiol 1988; 41:1151-1156.

10

Yarnell JW, Beswick AD, Sweetnam PM, Riad-Fahmy D. Endogenous sex hormones and ischemic heart disease in men. The Caerphilly prospective study. Arterioscler Thromb 1993; 13:517-520.

11

Joosten E, van Den Berg A, Riezler R et al. Metabolic evidence that deficiencies of vitamin B-12 (cobalamin), folate, and vitamin B-6 occur commonly in elderly people. Am J Clin Nutr 1993; 58:468-476.

12

Selhub J, Jacques PF, Wilson PW, Rush D, Rosenberg IH. Vitamin status and intake as primary determinants of homocysteinemia in an elderly population. JAMA 1993; 270:2693-2698.

13

Nygard O, Vollset SE, Refsum H, Brattstrom L, Ueland PM. Total homocysteine and cardiovascular disease. Journal of Internal Medicine 1999; 246;425-454.

14

Eikelboom JW, Lonn E, Genest JJr, Hankey G, Yusuf S. Homocyst(e)ine and cardiovascular disease: a critical review of the epidemiologic evidence. Ann Intern Med 1999; 131:363-375.

15

Bonaa KH, Njolstad I, Ueland PM et al. Homocysteine lowering and cardiovascular events after acute myocardial infarction. N Engl J Med 2006; 354:1578-1588.

16

The Heart Outcomes Prevention Evaluation I. Homocysteine lowering with folic acid and B vitamins in vascular disease. N Engl J Med 2006; 354:1567-1577.

17

Fonseca V, Guba SC, Fink LM. Hyperhomocysteinemia and the endocrine system: implications for atherosclerosis and thrombosis. Endocr Rev 1999; 20:738-759.

18

Andersson A, Hultberg B, Brattstrom L. Decreased serum homocysteine in pregnancy. Eur J Clin Chem Clin Biochem 1992; 30:377-379.

19

Yarali H, Yildirir A, Aybar F et al. Diastolic dysfunction and increased serum homocysteine concentrations may contribute to increased cardiovascular risk in patients with polycystic ovary syndrome. Fertil Steril 2001; 76:511-516.

20

Giltay EJ, Hoogeveen EK, Elbers JMH, Gooren LJG, Asscheman H, Stehouwer CDA. Effects of sex steroids on plasma total homocysteine levels: a study in transsexual males and females. J Clin Endocrinol Metab 1998; 83:550553.

21

Ebenbichler CF, Kaser S, Bodner J et al. Hyperhomocysteinemia in bodybuilders taking anabolic steroids. Eur J Intern Med 2001; 12:43-47.

26

Androgens and homocysteine 22

Yesilova Z, Ozata M, Oktenli C, Sanisoglu SY, Erbil MK, Dagalp K. Effect of supraphysiologic doses of testosterone on fasting plasma total homocysteine concentrations in men with Klinefelter's syndrome. Fertil Steril 2004; 81: 12781282.

23

Muller M, Aleman A, Grobbee DE, De Haan EHF, van Der Schouw YT. Endogenous sex hormone levels and cognitive function in aging men: Is there an optimal level? Neurology 2005; 64:866-871.

24

Aleman A, Muller M, De Haan EHF, van Der Schouw YT. Vascular risk factors and cognitive function in a sample of independently living men. Neurobiology of Aging 2005; 26:485-490.

25

Rasmussen K, Moller J, Lyngbak M. Within-person variation of plasma homocysteine and effects of posture and tourniquet application. Clin Chem 1999; 45:1850-1855.

26

Vermeulen A, Verdonck L, Kaufman JM. A critical evaluation of simple methods for the estimation of free testosterone in serum. J Clin Endocrinol Metab 1999; 84:3666-3672.

27

Sodergard R, Backstrom T, Shanbhag V, Carstensen H. Calculation of free and bound fractions of testosterone and estradiol-17 beta to human plasma proteins at body temperature. J Steroid Biochem 1982; 16:801-810.

28

Ocke M, Bueno-De-Mesquita H, Pols M, Smit H, Van Staveren W, Kromhout D. The Dutch EPIC food frequency questionnaire. II. Relative validity and reproducibility for nutrients. Int. J. Epidemiol. 1997 26 49S-58.

29

Stolk RP, Wink O, Zelissen PM, Meijer R, van Gils AP, Grobbee DE. Validity and reproducibility of ultrasonography for the measurement of intra-abdominal adipose tissue. Int J Obes Relat Metab Disord 2001; 25:1346-1351.

30

Tornaghi G, Raiteri R, Pozzato C et al. Anthropometric or ultrasonic measurements in assessment of visceral fat? A comparative study. Int J Obes Relat Metab Disord 1994; 18:771-775.

31

Voorrips LE, Ravelli AC, Dongelmans PC, Deurenberg P, Van Staveren WA. A physical activity questionnaire for the elderly. Med Sci Sports Exerc 1991; 23:974-979.

32

Cockcroft DW, Gault MH. Prediction of creatinine clearance from serum creatinine. Nephron 1976; 16:31-41.

33

Zmuda JM, Bausserman LL, Maceroni D, Thompson PD. The effect of supraphysiologic doses of testosterone on fasting total homocysteine levels in normal men. Atherosclerosis 1997; 130:199-202.

34

Oktenli C, Yesilova Z, Ozata M et al. Gonadotropin treatment increases homocysteine levels in idiopathic hypogonadotropic hypogonadism: an indirect effect mediated by changes in body composition. J Endocrinol 2003; 179:35-39.

35

Ueland PM, Refsum H. Plasma homocysteine, a risk factor for vascular disease: plasma levels in health, disease, and drug therapy. J Lab Clin Med 1989; 114:473-501.

36

Giri S, Thompson PD, Taxel P. Oral estrogen improves serum lipids, homocysteine and fibrinolysis in elderly men. Atherosclerosis 1998; 137:359-366.

37

Sudhir K, Komesaroff PA. Cardiovascular actions of estrogens in men. J Clin Endocrinol Metab 1999; 84:3411-3415.

27

Chapter 2 38

Sader MA, Mccredie RJ, Griffiths KA, Wishart SM, Handelsman DJ, Celermajer DS. Oestradiol improves arterial endothelial function in healthy men receiving testosterone. Clinical Endocrinology 2001; 54:175-181.

39

Komesaroff PA, Fullerton M, Esler MD, Dart A, Jennings G, Sudhir K. Low-dose estrogen supplementation improves vascular function in hypogonadal men. Hypertension 2001; 38:1011-1016.

40

Lew R, Komesaroff P, Williams M, Dawood T, Sudhir K. Endogenous estrogens influence endothelial function in young men. Circ Res 2003; 93:1127-1133.

41

Dimitrova KR, Degroot KW, Pacquing AM et al. Estradiol prevents homocysteine-induced endothelial injury in male rats. Cardiovascular Research 2002; 53:589-596.

42

Blouin K, Despres J-P, Couillard C et al. Contribution of age and declining androgen levels to features of the metabolic syndrome in men. Metabolism 2005; 54:1034-1040.

43

Trivedi DP, Khaw KT. Dehydroepiandrosterone sulfate and mortality in elderly men and women. J Clin Endocrinol Metab 2001; 86:4171-4177.

44

Martina V, Benso A, Gigliardi VR et al. Short-term dehydroepiandrosterone treatment increases platelet cGMP production in elderly male subjects. Clinical Endocrinology 2006; 64:260-264.

45

Bostom AG, Jacques PF, Nadeau MR, Williams RR, Ellison RC, Selhub J. Post-methionine load hyperhomocysteinemia in persons with normal fasting total plasma homocysteine: initial results from The NHLBI Family Heart Study. Atherosclerosis 1995; 116:147-151.

28

Chapter 2.2. Androgens and chronic low-grade inflammation

Manuscript based on this chapter: Nakhai Pour HR, Grobbee DE, Muller M, van der Schouw YT. Association of endogenous sex hormone with C-reactive protein levels in middle-aged and elderly men. Clinical Endocrinology 2007; 66:394-398.

29

Chapter 2 Summary______________________________________ __________________________________ Background: In women, postmenopausal estrogen supplementation increases levels of systemic markers of inflammation, which are important predictors of coronary heart disease risk. Whether endogenous sex hormone levels in men are also related to systemic subclinical inflammation is still unknown. Subjects and methods: We tested the hypothesis that higher endogenous sex hormones levels within the physiologic range may be associated with systemic subclinical inflammation. Circulating sex hormone and highsensitivity C-reactive protein (hs-CRP) levels were determined in 400 apparently healthy men aged between 40 and 80 years. We used multivariate linear regression analysis with the various sex hormones as determinant, and natural log hs-CRP as outcome. Results: Higher levels of total as well as bioavailable estradiol were associated with increased natural log hsCRP levels, which remained statistically significant after adjustment for age and cardiovascular risk factors, Natural log hs-CRP was 0.26 mg/l higher (95% CI -0.02; 0.54;) in the fourth compared to the first quartile of total estradiol, the p-value for linear trend was 0.05. For bioavailable estradiol the difference in natural log hs-CRP between the fourth and the first quartile was 0.30 mg/l higher (95% CI 0.03; 0.56; p-value test for linear trend 0.04). After adjustment for age and cardiovascular risk factors, physiologic levels of total or bioavailable testosterone or DHEAS were not associated with hsCRP. Conclusion: In conclusion, endogenous total and bioavailable estradiol levels were significantly associated with CRP among middle-aged and elder men.

30

Androgens and chronic low-grade inflammation Introduction______________________________________________________________________ Males do not experience a menopause, but sex hormone levels do decrease gradually with advancing age 1. Male gender is considered an independent risk factor for cardiovascular disease, and aging in men is associated with the state of endogenous androgen deficiency and increased risk for ischemic heart disease and cerebrovascular disease 2-4. Low-grade chronic inflammation has emerged as an important independent predictor of future coronary heart disease (CHD) in healthy individuals and of CHD death in patients with unstable angina pectoris or non-Q-wave myocardial infarction 5,6. It has been shown that with aging inflammation may have an increasing independent effect on the risk of cardiovascular disease 7. Given the results of the recent studies on effect of hormone replacement in postmenopausal women on plasma concentrations of hs-CRP, the question whether in men sex hormone levels are also associated with low-grade chronic inflammation 8-12 is still actual. In normal elderly men, pharmaceutical suppressing of endogenous testosterone production seems to increase the inflammatory response 13. It is not known whether the progressive decline in testosterone as a result of aging affects inflammatory status. Therefore, in the present cross-sectional study, we sought to determine the association between the levels of different endogenous sex hormones and CRP among middle-aged and elderly men. Subjects and methods

.

Subjects HAMLET is a population-based cohort of men (n= 400) aged 40 to 80 years and apparently healthy at time of enrollment. The HAMLET-study was set up to study the association between endogenous sex hormone levels and chronic disease risk, among which cardiovascular disease risk, which was established looking at risk factors and intermediate endpoints. Details of the study design, recruitment, and procedures have been published 14. The study was approved by the institutional Review Board of the University Medical Center Utrecht and written informed consent was obtained from all participants.

31

Chapter 2 High sensitivity serum (hs-CRP) measurement High sensitivity C-reactive protein (hs-CRP) was measured using a Behring Nefelometer II (Dade Behring, Liederbach, Germany). The lower limit of detection was 0.175 mg/l and the inter-assay variation was 2.4%. hs-CRP levels >10 mg/l can be taken as evidence of active inflammatory processes (e.g., trauma, infection), therefore subjects with hs-CRP>10 mg/l were excluded (n=16). Androgenic hormone determinations Circulating androgens levels were measured in serum. Total testosterone (TT) was measured after diethyl ether extraction using an in-house competitive radioimmunoassay employing a polyclonal antitestosterone-antibody (Dr Pratt AZG 3290). The lower limit of detection was 0.24 nmol/l and interassay variation was 6.0; 5.4 and 8.6% at 2.1; 5.6 and 23 nmol/l respectively (n=85). Bioavailable testosterone (free plus albumin bound) was calculated using measures of total testosterone and immunoassayed SHBG concentrations 15. Estradiol (E2) was measured after diethylether extraction and Sephadex chromatography using an in-house competitive radioimmunoassay employing a polyclonal anti-estradiol-antibody (Dr F de Jong, Erasmus MC, Rotterdam NL). The lower limit of detection was 20 pmol/l respectively (n=24, resp.17). Bioavailable E2 was calculated using the method described by Södergard et al 16. This method is based on the knowledge of the total concentration of all steroids competing for the same binding site on SHBG, the concentration of albumin (using a fixed concentration of 40 gr/l), the binding capacity of SHBG, and the association constant of E2 to the binding proteins. Dehydroepiandrosterone-sulfate (DHEA-S) was measured using an immunometric technique on an advantage chemiluminescense system (Nichols Institute Diagnostics, San Juan Capistrano, USA). The lower limit of detection was 0.1 µ mol/l and inter-assay variation was 5.2; 5.6 and 4.2% at 1.0; 4.9 and 14.2 µ mol/l respectively (n=19). Sex hormone binding globulin (SHBG) was measured using an immunometric technique on an IMMULITE Analyzer (Diagnostic Products Corporation, Los Angeles, USA). The lower limit of detection was 5 nmol/l and inter-assay variation was 1; 4.9 and 6.9% at 11.6; 36 and 93 nmol/l respectively (n=30). Other measurements A medical doctor obtained information on the prevalence of disease and the use of medication from a specified medical history. Height, weight, waist and hip circumference were measured with the participant in standing position wearing indoor clothes and no shoes. Body mass index (BMI) was calculated as the weight in kilograms divided by the square of height in meters. Blood pressure was measured in the morning after 10 min rest, twice in the right brachial artery (lying position) with a semi-automated device (Dinamap 8100; Critikon Inc., Tampa, Finland). The average of the two measurements of systolic and diastolic blood pressure was used for analysis. Visceral and subcutaneous fat were assessed using ultra-sound measurement. Ultrasonography was performed with an HDI 3000 (Philips Medical Systems, Eindhoven, the Netherlands) using a C 4-2 transducer.

32

Androgens and chronic low-grade inflammation For all images, the transducer was placed on a straight line drawn between the left and right midpoints between the lower rib and iliac crest. Measurements were made at the end of quiet expiration, applying minimal pressure without displacement of intraabdominal contents as observed by ultrasound image. Visceral fat comprises the distance between peritoneum and lumbar spine. Subcutaneous fat comprises the distance between the skin and the linea alba. For the measurement of visceral fat mass, we measured distances from three different angles (medial, left and right). For the subcutaneous fat mass the distance from the medial angle was measured in threefold. The mean of the three different measurements was calculated and used in our analyses 17. Smoking was estimated from self-report and categorized in current, former, and never smokers. Diseases were classified according to the International Classification of Diseases, 10th version (ICD-10). Hypertension was defined as systolic blood pressure ≥ 160 mmHg and/or use of anti-hypertensive medication. Diabetes mellitus was defined as treatment with insulin or oral hypoglycemic agents and/or fasting plasma venous glucose > 6.9 mmol/l. Data analysis We examined whether continuous variables were normally distributed using the Kolmogorov-Smirnov test. The distribution of hs-CRP levels was highly skewed; therefore, hs-CRP values were transformed by calculating the natural logarithm. Linear regression analysis was used to assess the association between endogenous sex hormones and levels hs-CRP. Quartiles of serum androgens levels were used as dummy variables in the models. Regression models were adjusted for age, and additionally for SHBG, BMI, hypertension, diabetes mellitus, alcohol intake, physical activity, smoking, and intra-abdominal fat. A stratified analysis according to different age strata has been done and the results were presented when the interaction term was statistically significant (P-value < 0.05). The data were analyzed using the SPSS statistical package for Windows version 11.0 (SPSS Inc., Chicago, IL, USA). Results__________________________________________________________________________ Mean age of subjects was 60 years and the median hs-CRP-level was 1.2 (interquartile range 0.702.70) mg/l. Characteristics of the study population across quartiles of bioavailable estradiol levels are presented in table 1. Subjects with the highest bioavailable estradiol level were younger, more likely to have visceral fat. Median hs-CRP levels were 1.10, 1.15, 1.30, 1.40 mg/l in the first, second, third and the fourth quartile of total estradiol. For total testosterone, median hs-CRP levels were 1.50, 1.50, 1.40 and 0.80 mg/l in the first, second, third and fourth quartile, respectively.

33

34 24.7 8.2 22.7 Median(IQR) 1.2 (0.6-2.7)

Hypertension Diabetes mellitus Current smoking hs-CRP (mg/l) IQR: Interquartile range

63.3 ± 10.9 142.1 ± 24.0 81.1 ± 10.5 25.9 ± 3.0 6.9 ± 2.0 %

Mean±SD

Q1(8.13-34.95)

Age (year) Systolic blood pressure (mmHg) Diastolic blood pressure (mmHg) BMI (Kg/m2) Intra-abdominal fat (cm)

Bioavailable estradiol (pmol/l)

1.1 (0.7-2.1)

24.7 9.3 21.6 Median(IQR)

59.2 ± 11.5 142.0 ± 22.6 80.3 ± 10.3 26.5 ± 4.1 7.3 ± 2.2 %

Mean±SD

Q2(34.95-40.10)

1.2 (0.7-2.4)

23.0 16.1 18.4 Median(IQR)

60.5 ± 11.7 143.5 ± 22.2 81.1 ± 10.2 26.5 ± 3.5 7.8 ± 2.3 %

Mean±SD

Q3(40.10-48.23)

Table 1. Characteristics of subjects according to quartiles of total and bioavailable estradiol.

1.5 (0.7-3.0)

30.6 10.2 30.6 Median(IQR)

57.8 ± 10.5 145.0 ± 18.9 83.1 ± 10.4 26.3± 3.3 8.0 ± 2.1 %

Mean±SD

Q4(48.20-110.57)

0.14

0.61 0.31 0.23

0.01 0.77 0.34 0.60 0.00

P-value

Chapter 2

Androgens and chronic low-grade inflammation Crude and adjusted β coefficients and 95% CIs for the association between hs-CRP and quartiles of endogenous androgens are shown in table 2. The hs-CRP-concentration increased with increasing levels of total as well as bioavailable estradiol. Adjusted for age, increasing levels of both total and bioavailable estradiol were associated with higher hs-CRP-levels. These associations only slightly attenuated after additional adjustment for cardiovascular risk factors, among which BMI. Natural log hs-CRP levels were 0.02 (95% CI -0.22; 0.27), 0.14 (-0.14; 0.41) and 0.26 (95% CI -0.02; 0.54) mg/l higher in the second, third and the fourth compared to the first quartile for total estradiol (pvalue test for linear trend 0.05), respectively. Furthermore, natural log hs-CRP levels were 0.06 (95% CI -0.21; 0.32), 0.02 (-0.25; 0.29) and 0.30 (95% CI 0.03; 0.56) mg/l higher in the second, third and the fourth compared to the first quartile for bioavailable estradiol (p-value test for linear trend 0.04), respectively. An inverse association was observed between quartiles of total and bioavailable testosterone and hs-CRP. After adjustment for age, testosterone (total, bioavailable) and DHEAS levels were not statistically significantly associated with hs-CRP levels. Additional adjustment for BMI, diabetes mellitus, hypertension, and smoking further attenuated the associations.

35

36 Sex hormone β (95% CI) 1 β (95% CI ) 2 β (95% CI) 3 Total estradiol Q1 (20.00-74.99 pmol/l) (ref) (ref) (ref) Q2 (75.00-89.99 pmol/l) 0.02 (-0.24, 0.29) 0.02 (-0.23, 0.27) 0.02 (-0.22, 0.27) Q3 (90.00-104.99 pmol/l) 0.20 (-0.10, 0.49) 0.19 (-0.09, 0.47) 0.14 (-0.14; 0.41) Q4 (105.00-180 pmol/l) 0.28 (-0.02, -0.58) 0.32 (0.03, 0.61) 0.26 (-0.02, 0.54) P-value for trend 0.05 0.02 0.05 Bioavailable estradiol Q1 (8.13-35.23 pmol/l) (ref) (ref) (ref) Q2 (35.24-40.65 pmol/l) -0.04 (-0.32, 0.23) 0.08 (-0.20, 0.35) 0.06 (-0.21, 0.32) Q3 (40.66-48.24 pmol/l) -0.04 (-0.32, 0.24) 0.05 (-0.23, 0.32) 0.02 (-0.25, 0.29) Q4 (48.25-82.11 pmol/l) 0.21 (-0.07, 0.48) 0.36 (0.09, 0.63) 0.30 (0.03, 0.56) P-value for trend 0.14 0.02 0.04 Total testosterone Q1 (0.24-14.99 nmol/l) (ref) (ref) (ref) Q2 (15.00-18.39 nmol/l) -0.14 (-0.42, 0.14) -0.09 (-0.36, 0.19) -0.13 (-0.40, 0.13) Q3 (18.40-21.99 nmol/l) -0.24 (-0.54, 0.06) -0.07 (-0.38, 0.23) -0.09 (-0.39, 0.20) Q4 (22.00-39.60 nmol/l) -0.58 (-0.91, -0.25) -0.25 (-0.60, 0.11) -0.19 (-0.53, 0.15) P-value for trend 0.00 0.23 0.36 Bioavailable testosterone Q1 (0.08-6.74 nmol/l) (ref) (ref) (ref) Q2 (6.75-7.83 nmol/l) 0.02 (-0.26, 0.30) 0.07 (-0.21, 0.36) 0.12 (-0.16, 0.39) Q3 (7.84-9.36 nmol/l) -0.17 (-0.46; 0.12) -0.04 (-0.34, 0.26) 0.03 (-0.26; 0.32) Q4 (9.37-16.33 nmol/l) -0.55 (-0.84, -0.25) -0.29 (-0.61,-0.04) -0.23 (-0.56, 0.09) P-value for trend 0.00 0.07 0.12 DHEAS Q1 ( 0.20- 4.19 µ mol/l) (ref) (ref) (ref) Q2 (4.20 -6.29 µ mol/l) -0.07 (-0.35, 0.21) 0.07 (-0.21, 0.35) 0.90 (-0.17, 0.36) Q3 ( 6.30-8.79 µ mol/l) -0.15 (-0.42, 0.13) 0.11 (-0.18, 0.40) 0.11 (-0.17, 0.39) Q4 ( 8.80-21.00µ mol/l) -0.33 (-0.60, -0.06) 0.05 (-0.27, 0.36) 0.00 (-0.30, 0.31) P-value for trend 0.02 0.87 0.91 1: Crude model 2: Model adjusted for age 3: Model adjusted for age, SHBG, BMI, hypertension, diabetes mellitus, alcohol intake, physical activity and smoking. 4: Model 3 additionally adjusted for intra-abdominal fat

(ref) 0.18 (-0.08, 0.44) 0.17 (-0.10, 0.44) 0.08 (-0.22, -0.37) 0.72

(ref) 0.13 (-0.14, 0.39) 0.06 (-0.22, 0.34) -0.15 (-0.46, -0.16) 0.29

(ref) -0.05 (-0.31, 0.21) 0.01 (-0.27, 0.30) -0.08 (-0.42, 0.25) 0.78

(ref) -0.01 (-0.27, 0.24) -0.09 (-0.36, 0.17) 0.13 (-0.13, 0.39) 0.45

(ref) 0.03 (-0.21, 0.27) 0.08 (-0.19, 0.34) 0.14 (-0.14, 0.42) 0.28

β ( 95% CI )4

Table 2. Multiple linear regression adjusted mean change in natural log of hs-CRP ((β) coefficients and 95% confidence intervals) by endogenous sex hormone levels.

Chapter 2

Androgens and chronic low-grade inflammation Discussion_______________________________________________________________________ In this study we found that low-grade chronic inflammation was independently related to endogenous estradiol levels, total as well as bioavailable. To appreciate these finding, some issues need to be addressed. Total testosterone and estradiol were measured using a radioimmunoassay method, which is more reliable compared with commercially available direct assays 18. In this study, blood samples were obtained in between 08:00 and 10:00h, and in standardized condition, which is necessary to obtain reliable measurements due to the possible daily variation of sex hormones levels 19. Previous studies already suggested the important implication for central obesity on circulating levels of testosterone and estradiol and moreover, inflammatory markers such as hs-CRP. Therefore we performed analyses with additional adjustment for the amount of visceral fat as measured by ultrasound 20. The visceral adipose tissue is an important source of cytokines, and thereby a contributor to a pro-inflammatory milieu 21,22. The few existing studies that evaluated the association between endogenous sex hormones and inflammatory markers in men have yielded inconsistent results. The present study is consistent with previous study not finding an association between physiologic levels of endogenous testosterone and low-grade chronic inflammation 23. Our finding, at least in part, is similar to the results of study in an occupation-based cohort of 715 healthy middle-aged men that found no correlation between testosterone and estradiol with CRP 24. In contrast to this evidence, in a population-based cohort of non-diabetic men with mild hypoandrogenism an inverse association of endogenous testosterone levels with CRP has been observed together with other components of the metabolic syndrome 25. To our knowledge, this is the first study that reported a positive association of endogenous estrogen and hs-CRP levels in middle-aged and elderly men. It is consistent with the finding that estrogen treatment in middle-aged and elderly men with prostate cancer under hormonal therapy is associated with increased circulating CRP concentration 26. In contrast, a study in 715 healthy young and middle-aged men, has shown no correlation between serum total and free estradiol and CRP 24, although effect estimates were not given. Whereas testosterone is the major sex hormone in males, several studies have recently emphasized the role of estrogens in male physiology 27. Estradiol, the major biologically active estrogen, is a metabolite of testosterone and about 80% is formed in peripheral tissues from testosterone, in man. It has been shown already that the age-associated decrease in testosterone levels is scarcely reflected in plasma estradiol levels, as a result of increasing aromatase activity with age and the age-associated increase in fat mass 28,29. Although the mechanism of the relation between estrogen and inflammatory markers is still poorly understood, it has been speculated that the effect of estrogen on hs-CRP is most likely an effect on gene expression in the liver 30. Strong support for this assumption that this effect is mediated by estrogen itself and not by a factor stimulated by estrogen, comes from the observations that the effect is dose-dependent 31, is seen in both women and men, and occurs only after oral treatment. It is not clear whether the effect of estrogen on gene expression is mediated by transcriptional activation or RNA stabilization. No hormone response elements have been discovered in the promoter of the CRP gene, but one study

37

Chapter 2 has shown that estrogen can stimulate the transcription factor C/EBP-ß, which is involved in CRP transcription 32. In addition, recent studies have shown that adipose tissue is not only the main source of peripheral conversion of testosterone to estradiol, but has itself endocrine capacities and produces adipokines like adiponectin, and interleukin-6 (IL-6) 33,34, inducing chronic low-grade inflammation through increased CRP-production by the liver. In our data, adjustment for BMI did not affect our findings. However, we also had ultrasonographically measured amount of intra-abdominal fat available, and when we adjusted for intra-abdominal fat, the association of both total and bioavailable estradiol levels with hs-CRP were no longer statistically significant (Table 3). From our study, we cannot separate the effects of estradiol and adipokines. Although the associations we found were statistically significant, one could question whether they have clinical relevance, as we studied endogenous sex hormone levels in the physiological range. However, when transforming the beta coefficients for the differences between the fourth and the first quartile of back from the natural logs, this results in a mean increase in hsCRP of appr. 1.3 mg/dl from the first quartiles of total and bioavailable estradiol to the fourth, which is associated with an approximately 50% increase in cardiovascular disease risk 35. Although we cannot distinguish whether these CRP-increases are a result of increased conversion to estradiol in peripheral fat tissue, or an increased inflammatory reaction in the fat tissue, weight reduction will decrease CRP levels either way, which will result in decreased cardiovascular risk by more mechanisms than just lowering of CRP. In conclusion, higher endogenous total and bioavailable estradiol levels were significantly associated with higher CRP-levels among middle-aged and elder men. Although the exact mechanism cannot be determined, weight reduction will lead to lower CRP-levels and lower cardiovascular disease risk References__________________________________

__________________________________

1.

Lamberts SWJ, van den Beld AW, van der Lely A-J. The endocrinology of aging. Science 1997; 278:419-424.

2.

Thom TJ. International mortality from heart disease: rates and trends. Int J Epidemiol 1989; 18:20-28.

3.

Vermeulen A, Kaufman JM. Ageing of the hypothalamo-pituitary-testicular axis in men. Horm Res 1995; 43:25-28.

4.

English KM, Mandour O, Steeds R, Diver MJ, Jones TH, Channer KS. Men with coronary artery disease have lower levels of androgens than men with normal coronary angiograms. Eur Heart J 2000; 21:890-894.

5.

Ridker PM, Cushman M, Stampfer MJ, Tracy RP, Hennekens CH. Inflammation, Aspirin, and the risk of cardiovascular disease in apparently healthy men. N Engl J Med 1997; 336:973-979.

6.

Liuzzo G, Biasucci LM, Gallimore JR et al. The prognostic value of C-reactive protein and serum amyloid a protein in severe unstable angina. N Engl J Med 1994; 331:417-424.

38

Androgens and chronic low-grade inflammation 7.

Abbott RD, Curb JD, Rodriguez BL et al. Age-related changes in risk factor effects on the incidence of coronary heart disease. Annals of Epidemiology 2002; 12:173-181.

8.

Cushman, M, Meilahn EN, Psaty BM, Kuller LH, Dobs AS,Tracy RP. Hormone replacement therapy, inflammation and hemostasis in elderly women. Arterioscler Thromb Vasc Biol 1999; 19:893-899.

9.

Cushman M, Legault C, Barrett-Connor E et al. Effect of postmenopausal hormones on inflammation-sensitive proteins : the postmenopausal estrogen/progestin interventions (PEPI) study. Circulation 1999; 100:717-722.

10.

Hu P, Greendale GA, Palla SL et al. The effects of hormone therapy on the markers of inflammation and endothelial function and plasma matrix metalloproteinase-9 level in postmenopausal women: The postmenopausal estrogen progestin intervention (PEPI) trial. Atherosclerosis 2005; 185:347-52.

11.

Barnes JF, Farish E, Rankin M, Hart DM. Effects of two continuous hormone therapy regimens on C-reactive protein and homocysteine. Menopaus 2005; 12:92-8.

12.

Ridker PM, Hennekens CH, Rifai N., Buring JE, Manson JE. Hormone replacement therapy and increased plasma concentration of C-reactive protein. Circulation 1999; 100:713-716.

13.

Khosla S, Atkinson EJ, Dunstan CR, O'Fallon WM. Effect of estrogen versus testosterone on circulating osteoprotegerin and other cytokine levels in normal elderly men. J Clin Endocrinol Metab 2002; 87:1550-1554.

14.

Muller M, Grobbee DE, den Tonkelaa I, Lamberts SWJ, van der Schouw YT. Endogenous sex hormones and metabolic syndrome in aging men. J Clin Endocrinol Metab 2005; 90:2618-2623.

15.

Vermeulen A., Verdonck L, Kaufman JM. A critical evaluation of simple methods for the estimation of free testosterone in serum. J Clin Endocrinol Metab 1999; 84:3666-3672.

16.


17.

Stolk RP, Wink O, Zelissen PM, Meijer R, van Gils AP, Grobbee DE. Validity and reproducibility of ultrasonography for the measurement of intra-abdominal adipose tissue. Int J Obes Relat Metab Disord 2001; 25:1346-1351.

18.

Rinaldi S, Dechaud H, Biess C et al. Reliability and validity of commercially available, direct radioimmunoassays for measurement of blood androgens and estrogens in postmenopausal women. Cancer Epidemiol Biomarkers Prev 2001; 10:757-765.

19.

Bremner WJ, Vitiello MV, Prinz PN. Loss of circadian rhythmicity in blood testosterone levels with aging in normal men. J Clin Endocrinol Metab 1983; 56:1278-81.

20.

Tanko LB, Bruun JM, Alexandersen P et al. Novel associations between bioavailable estradiol and adipokines in elderly women with different phenotypes of obesity: implications for atherogenesis. Circulation 2004; 110:2246-2252.

21.

Park HS, Park JY, Yu R. Relationship of obesity and visceral adiposity with serum concentrations of CRP, TNF[alpha] and IL-6. Diabetes Research and Clinical Practice 2005; 69:29-35.

22.

Piche M-E, Lemieux S, Weisnagel SJ, Corneau L, Nadeau A, Bergeron J. Relation of high-sensitivity C-reactive protein, interleukin-6, tumor necrosis factor-alpha, and fibrinogen to abdominal adipose tissue, blood pressure, and

39

Chapter 2 cholesterol and triglyceride levels in healthy postmenopausal women. The American Journal of Cardiology 2005; 96:92-97. 23.

Maggio M, Basaria S, Ble A et al. Correlation between testosterone and the inflammatory marker soluble interleukin-6 receptor in older men. J Clin Endocrinol Metab 2006; 91:345-347.

24.

Van Pottelbergh I, Braeckman L, De Bacquer D, De Backer G, Kaufman JM. Differential contribution of testosterone and estradiol in the determination of cholesterol and lipoprotein profile in healthy middle-aged men. Atherosclerosis 2003; 166:95-102.

25.

Laaksonen DE, Niskanen L, Punnonen K et al. Sex hormones, inflammation and the metabolic syndrome: a population-based study. Eur J Endocrinol 2003; 149: 601-608.

26.

Kovacs A, Henriksson P, Hamsten A, Wallen H, Bjorkegren J, Tornvall P. Hormonal regulation of circulating Creactive protein in men. Clin Chem 2005; 51:911-913.

27.

MacDonald PC, Madden JD, Brenner PF, Wilson JD, Siiteri PK. Origin of estrogen in normal men and in women with testicular feminization. J Clin Endocrinol Metab 1979; 49:905-916.

28.

Vermeulen A, Kaufman JM, Goemaere S, van Pottelberg I. Estradiol in elderly men. Aging Male 2002; 5:98-102.

29.

Cohen PG. Aromatase, adiposity, aging and disease. The hypogonadal-metabolic-atherogenic-disease and aging connection. Medical Hypotheses 2001; 56:702-708.

30.

Vongpatanasin W, Tuncel M, Wang Z, Arbique D, Mehrad B, Jialal I. Differential effects of oral versus transdermal estrogen replacement therapy on C-reactive protein in postmenopausal women. Journal of the American College of Cardiology 2003; 41:1358-1363.

31.

Prestwood KM, Unson C, Kulldorf, M, Cushman M. The effect of different doses of micronized 17{beta}-estradiol on C-reactive protein, interleukin-6, and lipids in older women. J Gerontol A Biol Sci Med Sci 2004; 59:827-832.

32.

Kousteni S, Han L, Chen J-R et al. Kinase-mediated regulation of common transcription factors accounts for the bone-protective effects of sex steroids. J. Clin. Invest 2003; 111:1651-1664.

33.

Trayhurn P, Wood ISA. Dipokines: inflammation and the pleiotropic role of white adipose tissue. Br J Nutr 2004; 92:347-55.

34.

Bruun JM, Lihn AS, Verdich C et al. Regulation of adiponectin by adipose tissue-derived cytokines: in vivo and in vitro investigations in humans. Am J Physiol Endocrinol Metab 2003; 285:527-533.

35.

Danesh J, Wheeler JG, Hirschfield GM et al. C-reactive protein and other circulating markers of inflammation in the prediction of coronary heart disease. N Engl J Med 2004; 350:1387-1397.

40

Chapter 3. Large artery stiffness and selected cardiovascular risk factors

41

42

Chapter 3.1. Large artery stiffness and endogenous androgens.

Manuscript based on this chapter: Nakhai Pour HR, Grobbee DE, Bots ML, Verhaar HJJ, Muller M, van der Schouw YT. Circulating sex hormone levels and aortic stiffness in men. Journal of the American Geriatrics Society 2007; 55:621-622. 43

Chapter 3 Summary________________________________________________________________________ Background: Recent evidence has suggested a protective role of endogenous sex hormones in the development of atherosclerosis. Aortic artery stiffening with aging results in hemodynamic alterations that have been associated with atherosclerosis and cardiovascular events. Sex hormone levels may be related to aortic stiffness. The present study was conducted to assess the association of physiologic levels of sex hormones and aortic stiffness in middle-aged and elder men. Subjects and methods: We conducted a population-based cross-sectional study among 376 men, aged 40-80. Endogenous sex hormone levels were measured using a radioimmunoassay with extraction methods. Aortic stiffness was assessed by measurement of carotid-femoral pulse wave velocity (PWV) using the applanation tonometry technique. Using multiple linear regression we compared mean PWV across quartiles of sex hormones, adjusted for age, mean arterial pressure, heart rate, and other potential confounders. Results: Although the unadjusted levels of several sex hormones were associated with aortic PWV, these associations disappeared after adjusting for age, mean arterial pressure, heart rate, and potential confounders. PWV was 0.14 (95% CI -0.52; 0.80), 0.04 (-0.54; 0.61) and 0.25 (-0.33; 0.82) m/s in the last quartile compared to the first quartile for testosterone (total, bioavailable) and DHEAS, respectively. Conclusion: Overall, in the present study, we found no independent relation between physiological levels of sex hormones and aortic stiffness over and above conventional cardiovascular risk factors.

44

Arterial stiffness and androgens Introduction______________________________________________________________________ The burden of atherosclerosis especially afflicts the older segment of the population. Large-artery stiffening has been recognized as an independent risk factor for cardiovascular disease, which increases with age 1. Aortic artery stiffening with aging results in hemodynamic alterations that have been associated with atherosclerosis and cardiovascular events 2-4. The underlying mechanisms of arterial stiffening are not completely understood. Recent evidence has emphasized a protective role of endogenous sex hormones in developments of atherosclerosis 5,6,7. It has been suggested that also androgenic hormones may have an independent effect on arterial stiffness. However, whether changes in endogenous sex hormones levels in men have an impact on arterial stiffness is not known. Androgen deprivation, which is widely used as treatment for prostate cancer, results in an increase in arterial stiffness 8,9. Studies on effects of endogenous androgens are scarce, but suggest that low testosterone levels modulate large artery stiffness in men 10,11. We set out to study the relation between circulating sex hormones and aortic stiffness, as measured by carotid-femoral PWV, in independently living men. Subjects and methods__________________________________ ___________________________ HAMLET is a population-based cohort of men (n= 400) aged 40 to 80 years and apparently healthy at time of enrollment. Details of the study design, recruitment, and procedures have been published previously 12. The study was approved by the institutional Review Board of the University Medical Center Utrecht and written informed consent was obtained from all participants. Androgenic hormone determinations A venapuncture was performed between 08:00h and 10:00h and a fasting blood sample was obtained. Platelet free serum was obtained by centrifugation and immediately stored at -20°C. Circulating androgens levels were measured in serum. Total testosterone (TT) was measured after diethyl ether extraction using an in-house competitive radioimmunoassay employing a polyclonal antitestosterone-antibody (Dr Pratt AZG 3290). The lower limit of detection was 0.24 nmol/l and interassay variation was 6.0; 5.4 and 8.6% at 2.1; 5.6 and 23 nmol/l respectively (n=85). Bioavailable testosterone (free plus albumin bound) was calculated using measures of total testosterone and immunoassayed SHBG concentrations 13. Estradiol (E2) was measured after diethylether extraction and Sephadex chromatography using an in-house competitive radioimmunoassay employing a polyclonal anti-estradiol-antibody (Dr FH de Jong, Erasmus MC, Rotterdam, The Netherlands). The lower limit of detection was 20 pmol/l respectively (n=24, resp.17). Bioavailable E2 was calculated using the method described by Södergard et al 14. This method is based on the knowledge of the total concentration of all steroids competing for the same binding site on SHBG, the concentration of albumin (using a fixed concentration of 40 gr/l), the binding capacity of SHBG, and the association constant of E2 to the binding proteins. Dehydroepiandrosterone-sulfate (DHEA-S) was measured

45

Chapter 3 using an immunometric technique on an advantage chemiluminescense system (Nichols Institute Diagnostics, San Juan Capistrano, USA). The lower limit of detection was 0.1 µ mol/l and inter-assay variation was 5.2; 5.6 and 4.2 % at 1.0; 4.9 and 14.2 µ mol/l respectively (n=19). Sex hormone binding globulin (SHBG) was measured using an immunometric technique on an IMMULITE Analyser (Diagnostic Products Corporation, Los Angeles, USA). The lower limit of detection was 5 nmol/l and inter-assay variation was 1; 4.9 and 6.9 % at 11.6; 36 and 93 nmol/l respectively (n=30). Pulse Wave velocity (PWV) measurement The PWV was determined non-invasively with use of SphygmoCor device (PWV Medical, Sydney, Australia), which allowed an online pulse wave recording and automatic calculation of PWV with two transducers (Millar SPT 301 pressure transducer; Milar Instruments, Sydney, Australia). Details of the procedure for aortic PWV measurement have been published previously 15. The average of 10 successive measurements was used in the analyses to cover a complete respiratory cycle. The whole procedure was repeated 3 times per subjects and the average PWV-value was used for the analysis. In order to evaluate the reproducibility of the technique in our laboratory, a subset of 20 participants had their PWV re-measured several weeks after their first visit. Absolute mean difference in PWV of the repeated measurements between visits was 0.09 m/s (SD 1.5 m/s). Data on carotidfemoral PWV measurement was available for 378 men. Other measurements A trained physician obtained information on prevalent diseases, medication use, and life-style factors. Participants were asked about current use of medications and examining labels of drugs brought to the clinic checked these reports. Smoking was estimated from self-report and categorized in current, former, and never smokers. Alcohol consumption was assessed by a validated Food Frequency Questionnaire (FFQ), designed to estimate regular intake of 178 food items in the year prior to enrolment 16. Height and weight were measured in standing position without shoes. Waist and hip circumference were measured using a standard household centimeter, and waist-to-hip ratio was calculated by dividing waist by hip circumference. Blood pressure (BP) was measured twice at the right brachial artery with a semi-automated device (Dinamap, GE HealthCare, Tampa, FL). The average of the two measurements of systolic and diastolic blood pressure was used for analysis and further calculation. Hypertension was defined as systolic blood pressure≥160 mmHg and/or diastolic blood pressure ≥ 100 mmHg and/or use of anti-hypertensive medication. Physical activity was assessed using a questionnaire validated in an elderly population 17. Ultrasonography was performed with an HDI 3000 (Philips Medical Systems, Eindhoven, the Netherlands) using a C 4-2 transducer. For all images, the transducer was placed on a straight line drawn between the left and right midpoints between the lower rib and iliac crest. Measurements were made at the end of quiet expiration, applying minimal pressure without displacement of intraabdominal contents as observed by ultrasound image. Visceral fat comprises the distance between peritoneum and lumbar spine.

46

Arterial stiffness and androgens Subcutaneous fat comprises the distance between the skin and the linea alba. For the measurement of visceral fat mass, we measured distances from three different angles (medial, left and right). For the subcutaneous fat mass the distance from the medial angle was measured in threefold. The mean of the three different measurements was calculated and used in our analyses 18. The subjects’ mean daily alcohol intake was estimated from a validated Food Frequency Questionnaire 16. Diabetes mellitus was defined as treatment with insulin or oral hypoglycaemic agents or fasting plasma venous glucose > 6.9 mmol/l. Presence of cardiovascular disease was defined as coronary heart disease, peripheral artery disease or stroke. Data analysis Data analysis was based on 376 subjects as two extreme values in PWV (2.75 and 30.51 m/s) have been excluded, since they were considered biologically not plausible. Distribution of anthropometric, lifestyle characteristic, plasma sex hormone levels, and cardiovascular risk factors were expressed as mean (SD) for continuous normally distributed variables and numbers (%) for categorical variables. P≤ 0.05 was considered statistically significant. We used multiple linear regression analysis to estimate mean difference in carotid-femoral PWV between higher quartiles of sex hormones and the first, adjusting for potential confounding variables. We constructed 3 sets of regression models. In the first set (model 1), adjustment has been done for age. Then in the second set of models (model 2) we included variables that accounted for much of the variance in PWV: mean arterial pressure and heart rate. The third set of models also included the classical cardiovascular risk factors: waist-to-hip ratio, prevalent CVD, hypertension, diabetes mellitus, smoking status, daily alcohol intake, total cholesterol level, triglyceride level, and visceral fat. We conducted similar regression analyses for levels of the other hormones. Data analyses were performed using SPSS statistical software (version 12; SPSS, Chicago, IL, USA). Results__________________________________________________________________________ General characteristics of the study population are given in table 1. The population characteristics according to bioavailable testosterone and estradiol, and DHEAS quartiles are presented in table 2. Compared with men in lowest sex hormones quartiles, men with higher levels of all hormones appeared somewhat younger, had a lower prevalence of cardiovascular disease, hypertension and diabetes mellitus and were more likely to be current smokers. We found an unadjusted inverse association between testosterone (total, bioavailable) as well as DHEAS levels and PWV; PWV was 1.18 (95% CI -1.90; -0.45), 2.39 (-3.05; -1.72) and 1.94 (2.62; -1.25) m/s lower in the fourth quartile compared to the first for total testosterone, bioavailable testosterone, and DHEAS, respectively. Overall, after adjustment for age, the estimates attenuated and we did not observe a statistically significant association. In the fully adjusted model also no associations were observed; PWV was 0.14 (95% CI -0.52; 0.80), 0.04 (-0.54; 0.61) and 0.25 (-0.33; 0.82) m/s in the fourth quartile compared to the first for total and bioavailable testosterone, and

47

Chapter 3 DHEAS, respectively. In our population, serum total and bioavailable estradiol levels were not significantly associated with PWV with or without adjustment for aforementioned covariates (Table 3).

Table 1 General characteristics of the study population (n=376).

Characteristic

Mean±SD

Age(years) Mean arterial pressure(mmHg) Heart rate(beat/min) Waist-to-hip ratio Mean visceral fat Alcohol intake(g) Total cholesterol(mmol/l) Triglycerides(mmol/l) Carotid-femoral PWV(m/s) Hormone levels Total testosterone(nmol/l) Bioavailable testosterone(nmol/L) Total estradiol(pmol/l) Bioactive estradiol(pmol/l) SHBG(nmol/l) DHEAS(µ mol/l)

59.9±11.3 128.8±16.1 70.0±13.0 0.97±0.10 7.6±2.2 20.3±21.6 5.8±1.0 1.6±1.4 9.4±2.5

Cardiovascular disease Hypertension Diabetes mellitus Current smoking

48

18.6±5.3 14.5±5.7 91.2±23.0 42.2±11.6 40.6±14.4 6.7±3.3 % 14.1 26.1 10.6 24.6

Arterial stiffness and androgens

Table 2. General characteristics of the study population across quartiles of bioavailable testosterone, DHEAS, and bioavailable estradiol.

Bioavailable Testosterone (nmol/l) Age(years) Mean arterial pressure(mmHg) Waist-to-hip ratio Visceral fat(cm) Cholesterol(mmol/l) Triglyceride(mmol/l) Alcohol intake(g/day) PWV(m/s) CVD Hypertension DM Current smoking DHEAS(µmol/l) Age(years) Mean arterial pressure(mmHg) Waist-to-hip ratio Visceral fat(cm) Cholesterol(mmol/l) Triglyceride(mmol/l) Alcohol intake(g/day) PWV(m/s) CVD Hypertension DM Current smoking Bioavailable estradiol (pmol/l) Age(years) Mean arterial pressure(mmHg) Waist-to-hip ratio Visceral fat(cm) Cholesterol(mmol/l) Triglyceride(mmol/l) Alcohol intake(g/day) PWV(m/s) CVD Hypertension DM Current smoking

Q1 Mean±SD 9.5±3.1 66.4±8.8 132.0±16.71 0.98±0.05 7.3±2.16 5.5±1.05 1.5±1.25 18.5±22.32 10.5±2.56 % 20.2 41.5 18.1 18.1 Mean±SD 3.0±0.9 68.1±7.5 130.5±16.6 0.99±0.05 7.7±2.3 5.6±0.9 1.4±0.8 18.7±18.4 10.5±2.6 % 18.2 34.8 15.2 15.2 Mean±SD 30.0±4.8 63.5±11.0 128.6±17.2 0.96±0.06 7.3±2.1 5.6±1.0 1.4±1.2 19.1±17.2 9.7±2.4 % 14.7 26.3 7.4 25.3

Q2 Mean±SD 12.7±3.3 63.1±10.7 129.1±16.4 0.98±0.04 7.5±2.3 5.8±0.9 1.8±1.5 23.0±26.8 9.9±2.6 % 15.4 28.6 11 25.3 Mean±SD 5.1±0.6 62.3±10.2 130.8±15.3 0.97±0.05 7.5±2.3 5.8±0.9 1.5±1.1 16.0±17.6 9.6±2.4 % 13.0 23.9 13 19.6 Mean±SD 37.6±1.6 59.5±11.8 127.8±16.3 0.97±0.05 7.8±2.3 5.7±0.9 1.6±1.7 18.8±21.4 9.3±2.7 % 12.8 26.6 8.5 20.2



49

Chapter 3 Table 3 Multiple linear regression β-coefficients and 95% confidence intervals describing the differences in PWV (m/s) for the higher quartiles of hormone levels compared to lowest quartile.

Total Testosterone PWV(m/s) Crude model Adjusted model* Adjusted model† Adjusted model‡ Bioavalable Testosterone PWV(m/s) Crude model Adjusted model* Adjusted model† Adjusted model‡ Total Estradiol PWV(m/s) Crude model Adjusted model* Adjusted model† Adjusted model‡ Bioavailable Estradiol PWV(m/s) Crude model Adjusted model* Adjusted model† Adjusted model‡ DHEAS PWV(m/s) Crude model Adjusted model* Adjusted model† Adjusted model‡ SHBG PWV(m/s) Crude model Adjusted model* Adjusted model† Adjusted model‡

Q2 Mean ±SD 9.8±2.8 -0.01(-0.73; 0.70) -0.18(-0.75; 0.39) -0.11(-0.61; 0.40) 0.03(-0.48; 0.54)

Q3 Mean ±SD 9.3±2.2 -0.60(-1.31;0.10) -0.45(-1.01; 0.11) -0.20(-0.69; 0.30) 0.03(-0.52; 0.57)

Q4 Mean ±SD 8.7±2.2 -1.18(-1.90; -0.45) -0.51(-1.10; -0.07) -0.17(-0.69; 0.35) 0.14(-0.52; 0.80)

9.9 ± 2.6 -0.62(-1.30; 0.05) -0.20(-0.77; 0.37) -0.08(-0.58; 0.43) 0.09(-0.42; 0.59)

9.2 ± 2.2 -1.28(-1.96; -0.61) -0.37(-0.95; 0.21) -0.24(-0.75; 0.27) -0.10(-0.62; 0.42)

8.1 ± 1.9 -2.39(-3.05; -1.72) -0.40(-1.04; 0.24) -0.29(-0.85; 0.27) 0.04(-0.54; 0.61)

9.2 ± 2.4 -0.07(-0.78; 0.65) -0.21(-0.77; 0.35) -0.06(-0.56; 0.44) -0.11(-0.62; 0.39)

9.9 ± 2.7 0.66(-0.05; 1.37) 0.32(-0.24; 0.88) 0.25(-0.24; 0.75) 0.14(-0.36; 0.64)

9.4 ± 2.3 0.12(-0.57; 0.81) 0.13(-0.41; 0.68) 0.11(-0.37; 0.59) -0.02(-0.52; 0.48)

9.3 ± 2.7 -0.39(-1.11; 0.32) 0.17(-0.39; 0.74) 0.21( -0.28; 0.71) 0.15(-0.35; 0.65)

9.3 ± 2.2 -0.43(-1.15; 0.29) 0.17(-0.39; 0.74) 0.14( -0.36; 0.64) 0.04(-0.49; 0.56)

9.4 ± 2.7 -0.30(-1.01; 0.41) 0.56(-0.01; 1.12) 0.27( -0.23; 0.77) 0.06(-0.47; 0.59)

9.6 ± 2.4 -0.91(-1.61; -0.22) -0.08(-0.66; 0.50) -0.19(-0.70; 0.32) -0.11(-0.61; 0.40)

9.1 ± 2.1 -1.43(-2.12; -0.73) 0.11(-0.49; 0.72) 0.06(-0.47; 0.60) 0.07(-0.47; 0.60)

8.6 ± 2.4 -1.94(-2.62; -1.25) 0.31(-0.34; 0.95) 0.16(-0.40; 0.73) 0.25(-0.33; 0.82)

9.7 ± 2.7 1.16(0.46; 1.87) 0.47(-0.90; 1.03) 0.38(-0.12; 0.88) 0.50(-0.01; 1.00)

9.9 ± 2.4 1.39(0.68; 2.11) 0.19(-0.39; 0.78) 0.43(-0.09; 0.95) 0.58(0.06; 1.10)

9.5 ± 2.4 0.93(0.22; 1.64) -0.51(-1.10; 0.08) -0.10(-0.63; 0.43) 0.12(-0.44; 0.68)

* Model adjusted for age † Model adjusted for age, mean arterial pressure, heart rate ‡ Model adjusted for age, mean arterial pressure, heart rate, SHBG, hypertension, DM, CVD, waist-to-hip ratio, smoking, daily alcohol intake, total cholesterol, triglyceride and visceral fat

50

Arterial stiffness and androgens Discussion_______________________________________________________________________ The present study in of 376 middle-aged and elder men indicates that endogenous sex hormone levels within the physiologic range are not likely to be related to aortic stiffness, a marker of future cardiovascular disease. Some methodological aspects need to be addressed. Over the recent years measurement of arterial stiffness became an increasingly important part of the process of cardiovascular risk assessment. We used carotid-femoral PWV as a marker of aortic arterial wall stiffness, because PWV is a highly reproducible and non-invasive index of aortic rigidity 19. In this study, blood samples were obtained between 08:00h and 10:00h, which is necessary to obtain reliable measurements due to the possible daily variation of androgens 20. Endogenous sex hormone levels were measured using a radioimmunoassay with extraction method, which is more reliable compared with commercially available direct assays 21. However, the interpretability of the results still may be restricted by several factors inherent to the cross-sectional design. Recent studies examining arterial stiffness mostly focused on relation of arterial stiffness and non-physiological levels of sex hormone. An inverse association between serum testosterone levels and arterial stiffness has been shown in androgen-deprived prostate cancer patients and in male-to-female transsexuals 9,22. It has been shown that males with prostate cancer, rendered hypogonadal by LHRH analog therapy, experienced an increased of central arterial stiffness. The studies regarding endogenous physiological levels of sex hormones yielded contradictory results. In a study among 36 hypogonadal men and 25 eugonadal controls, the hypogonadal men had better endothelial dependent dilation than the eugonadal men at baseline, but only in the hypogonadal men lower testosterone levels were associated with better endothelial dependent dilation 23, suggesting that lower testosterone levels are related to decreased arterial stiffness. In a cross-sectional study among 55 middle-age men a higher free testosterone index was related to increased carotid-femoral PWV, i.e. stiffer arteries, was reported, but only for the subgroup of men not using antihypertensive medication 10, But we could not replicate this (data not shown). A large cross-sectional study among 508 male aged 41-72 did not find any associations between testosterone and arterial stiffness 24. Conversely, a recently published cross-sectional study among 206 men showed that lower endogenous T levels are associated with higher carotid artery stiffness 11. Together our finding and the current evidence suggest that testosterone levels within the normal range may do not have a major role in regulating vascular endothelial function as a major detrimental component of arterial stiffness. With respect to estrogen, this has been shown to enhance endothelium-dependent vasodilatation in premenopausal and postmenopausal women and to reduce arterial stiffness 25,26. Moreover, it has been suggested that endogenous estrogens may play a protective role in the male vascular system 27. However in male-to-female transsexuals, treatment with estrogen has consistently shown an improvement in endothelial function, as measured by flow-mediated dilation, but no effect on systemic arterial compliance was found. A recent placebo-controlled double-blind randomized trial in 20 healthy men also did not find an effect of estrogen level changes in the physiological range with

51

Chapter 3 large artery stiffness 28. Taking together with the findings of our study, estrogen does not seem to affect large artery stiffness in men. There is evidence from experimental and clinical studies, showing that DHEAS administration has beneficial effects on markers of vascular risk 29,30, angiographic evidence of atherosclerosis 31 and vascular function 32,33 however, the majority of epidemiological studies show that subjects with cardiovascular risk factors 34,35,36 cardiovascular morbidity 37,38 or radiological evidence of atherosclerotic vascular disease 39 have lower plasma levels of DHEAS. Our finding in present study is consistent with previous studies which show null association between DHEAS and arterial stiffness after adjustment for confounders. Apparently, in eugonadal men, other traditional cardiovascular risk factors are more important determinants of vascular stiffness than sex hormone levels in the physiologic range. In conclusion, our results from this cross-sectional study of 376 middle-age and elderly men suggest no independent association between endogenous androgens levels and aortic stiffness. References__________________________________

__________________________________

1.

Arnett DK, Evans GW, Riley WA. Arterial stiffness: a new cardiovascular risk factor? Am J Epidemiol 1994; 140:669682.

2.

Salomaa V, Riley W, Kark JD, Nardo C, Folsom A. Non–insulin-dependent diabetes mellitus and fasting glucose and insulin concentrations are associated with arterial stiffness indexes: the ARIC study. Circulation 1995; 91:1432-1443.

3.

Lehmann ED, Watts GF, Gosling RG. Aortic distensibility and hypercholesterolaemia. Lancet 1992; 340:1171-1172.

4.

van Popele NM, Grobbee DE, Bots ML et al. Association between arterial stiffness and atherosclerosis: the Rotterdam Study. Stroke 2001; 32:454-460.

5.

Hak AE, Witteman JCM, de Jong FH, Geerlings MI, Hofman A, Pols HAP. Low levels of endogenous androgens increase the risk of atherosclerosis in elderly men: the Rotterdam study. J Clin Endocrinol Metab 2002; 87:36323639.

6.

Muller M, van den Beld AW, Bots ML, Grobbee DE, Lamberts SWJ, van der Schouw YT. Endogenous sex hormones and progression of carotid atherosclerosis in elderly men. Circulation 2004; 109:2074-2079.

7.

van den Beld AW, Bots ML, Janssen JAMLL, Pols HAP, Lamberts SWJ, Grobbee DE. Endogenous hormones and carotid atherosclerosis in elderly men. Am. J. Epidemiol 2003; 157:25-31.

8.

Dockery F, Bulpitt CJ, Agarwal S, Rajkumar C. Testosterone suppression in men with prostate cancer is associated with increased arterial stiffness. Aging Male 2002; 5:216-222.

9.

Smith JC, Bennett S, Evans LM et al. The effects of induced hypogonadism on arterial stiffness, body composition, and metabolic parameters in males with prostate cancer. J Clin Endocrinol Metab 2001; 86:4261-4267.

10.

Dockery F, Bulpitt CJ, Donaldson M, Fernandez S, Rajkumar C. The relationship between androgens and arterial stiffness in older men. J Am Geriatr Soc 2003; 51:1627-1632.

52

Arterial stiffness and androgens 11.

Hougaku H, Fleg JL, Najjar SS et al. Relationship between androgenic hormones and arterial stiffness, based on longitudinal hormone measurements. Am J Physiol Endocrinol Metab 2006; 290:234-242.

12.

Muller M, Grobbee DE, den Tonkelaar I, Lamberts SWJ, van der Schouw YT. Endogenous sex hormones and metabolic syndrome in aging men. J Clin Endocrinol Metab 2005; 90:2618-2623.

13.

Vermeulen A, Verdonck L, Kaufman JM. A critical evaluation of simple methods for the estimation of free testosterone in serum. J Clin Endocrinol Metab 1999; 84:3666-3672.

14.


15.

Sierksma A, Muller M, van der Schouw YT, Grobbee DE, Hendriks HF, Bots ML. Alcohol consumption and arterial stiffness in men. J Hypertens 2004; 22:357-362.

16.

Ocke M, Bueno-de-Mesquita H, Pols M, Smit HA, van Staveren WA, Kromhout, D. The Dutch EPIC food frequency questionnaire. II. Relative validity and reproducibility for nutrients. Int. J. Epidemiol 1997; 26:49-58.

17.

Voorrips LE, Ravelli AC, Dongelmans PC, Deurenberg P, Van Staveren WA. A physical activity questionnaire for the elderly. Med Sci Sports Exerc 1991; 23:974-979.

18.

Stolk R, Wink O, Zelissen P, Meijer R, van Gils AP, Grobbee, DE. Validity and reproducibility of ultrasonography for the measurement of intra-abdominal adipose tissue. Int J Obes Relat Metab Disord 2001; 25:1346-1351.

19.

Asmar R, Benetos A, Topouchian J et al. Assessment of arterial distensibility by automatic pulse wave velocity measurement: validation and clinical application studies. Hypertension 1995; 26:485-490.

20.

Bremner WJ, Vitielo MN, Prinz PN. Loss of circadian rhythmicity in blood testosterone levels with aging in normal men. J Clin Endocrinol Metab 1983; 56:1278-1281.

21.

Thijssen JH. Laboratory tests in the endocrine evaluation of aging males. In: Lunenfeld B, Gooren LJ. Textbook of Men's Health. Parthenon Publishing Group. 2002;44-50.

22.

Giltay EJ, Lambert J, Gooren LJ, Elbers JMH, Steyn M, Stehouwer CDA. Sex steroids, insulin, and arterial stiffness in women and men. Hypertension 1999; 34:590-597.

23.

Zitzmann M, Brune M, Nieschlag E. Vascular reactivity in hypogonadal men is reduced by androgen substitution. J Clin Endocrinol Metab 2002; 87:5030-5037.

24.

Gyllenborg J, Rasmussen S, Broch-Johnsen K, Heitmann B, Skakkebaek N, Juul A. Cardiovascular risk factors in men: the role of gonadal steroids and sex hormone-binding globulin. Metabolism 2001; 50;882-8

25.

Gangar KF, Vyas S, Whitehead M, Crook D, Meire H, Campbell S. Pulsatility index in internal carotid artery in relation to transdermal oestradiol and time since menopause. Lancet 1991; 338:839-842.

26.

Hashimoto M, Akishita M, Eto M, Ishikawa M, Kozaki K, Toba K. Modulation of endothelium-dependent flow-mediated dilatation of the brachial artery by sex and menstrual cycle. Circulation 1995; 92:3431-3435.

53

Chapter 3 27.

Smith EP, Boyd J, Frank GR, Takahashi H, Cohen RM, Specker B. Estrogen resistance caused by a mutation in the estrogen-receptor gene in a man. N Engl J Med 1994; 331:1056-1061.

28.

Lew R, Komesaroff P, Williams M, Dawood T, Sudhir K. Endogenous estrogens influence endothelial function in young men. Circ Res 2003; 93:1127-1133.

29.

Beer N, Jakubowicz D, Matt D, Matt DW, Beer RM, Nestler JE. Dehydroepiandrosterone reduces plasma plasminogen activator inhibitor type 1 and tissue plasminogen activator antigen in men. Am J Med Sci 1996; 311:205210.

30.

Jesse R, Loesser K, Eich D, Qian YZ, Hess ML, Nestler JE. Dehydroepiandrosterone inhibits human platelet aggregation in vitro and in vivo. Ann N Y Acad Sci 1995; 29:281-290.

31.

Herrington DM. Dehydroepiandrosterone and coronary atherosclerosis. Ann N Y Acad Sci 1995; 29:271-280.

32.

Williams MRI, Dawood T, Ling S et al. Dehydroepiandrosterone increases endothelial cell proliferation in vitro and improves endothelial function in vivo by mechanisms independent of androgen and estrogen receptors. J Clin Endocrinol Metab 2004; 89:4708-4715.

33.

Williams MRI, Ling S, Dawood T et al. Dehydroepiandrosterone Inhibits Human Vascular Smooth Muscle Cell Proliferation Independent of ARs and ERs. J Clin Endocrinol Metab 2002; 87:176-181.

34.

Shono N, Kumagai S, Higaki Y et al. The relationships of testosterone, estradiol, dehydroepiandrosterone-sulfate and sex hormone-binding globulin to lipid and glucose metabolism in healthy men. J Atheroscler Thromb 1996; 3:45-51.

35.

Okamoto K. Relationship between dehydroepiandrosterone sulfate and serum lipid levels in Japanese men. J Epidemiol. 1996; 6:63-67.

36.

Abbasi A, Duthie EJ, Sheldahl L et al. Association of dehydroepiandrosterone sulfate, body composition, and physical fitness in independent community-dwelling older men and women. J Am Geriatr Soc 1998; 46:263-273.

37.

Trivedi DP, Khaw KT. Dehydroepiandrosterone sulfate and mortality in elderly men and women. J Clin Endocrinol Metab 2001; 86:4171-4177.

38.

Feldman HA, Johannes CB, Araujo AB, Mohr BA, Longcope C, McKinlay JB. Low Dehydroepiandrosterone and Ischemic Heart Disease in Middle-aged Men: Prospective Results from the Massachusetts Male Aging Study. Am. J. Epidemiol 2001; 153:79-89.

39.

Herrington DM, Gordon GB, Achuff SC et al. Plasma dehydroepiandrosterone and dehydroepiandrosterone sulfate in patients undergoing diagnostic coronary angiography. J Am Coll Cardiol 1990; 16:862-870.

54

Chapter 3.2. Large artery stiffness and chronic low-grade inflammation.

Manuscript based on this chapter: Nakhai Pour HR, Grobbee DE, Bots ML, Muller M, van der Schouw YT. C - reactive protein and aortic stiffness and wave reflection in middle-aged and elderly men from the community. Journal of Human Hypertension (In press)

55

Chapter 3 Summary________________________________________________________________________ Background: High plasma C-reactive protein (hs-CRP) levels and arterial stiffness are risk factors for cardiovascular diseases. Pulse wave velocity (PWV) and augmentation index (AIx) have been found to be elevated in patients with vascular inflammation, diabetes mellitus, hypertension, hypercholesterolemia, and in smokers. Subjects and methods: We investigated the relation of hs-CRP with aortic stiffness in 362 men. The levels of hs-CRP were measured using a nephelometric method. Aortic PWV and AIx were assessed from carotid-femoral segment and radial artery waveforms with the use of the SphygmoCor device. Results: In the crude model, aortic PWV increased significantly with increasing serum hs-CRP levels; PWV increased by 2.48 m/s (95% CI 1.58; 3.38) in the fifth compared to the first quintile of hs-CRP. In the adjusted model, the PWV increased by 0.84 m/s (95% CI 0.13; 1.55) in the fifth quintile compared to the first quintile (p-value for trend was 6.9 mmol/l. CVD was considered present if there was pre-existing coronary heart disease, peripheral artery disease or stroke. Carotid intima-media thickness (CIMT) was used to assess the extent of subclinical atherosclerosis. Ultrasonography of both the left and right carotid arteries was performed using a 7.5 MHz linear array transducer (Acuson Aspen, Mountain View, California, USA). Details of the procedure for carotid intima-media thickness measurement have been published previously 16.

58

Arterial stiffness and chronic low-grade inflammation Data analysis Of the 400 subjects who participated in the HAMLET study, aortic PWV and AIx were measured in 378 subjects. Missing information on both measures was entirely due to logistic reasons. Two extreme values in PWV (2.75 and 30.51 m/s) have been excluded, since they were considered biologically implausible. hs-CRP levels >10 mg/l can be taken as evidence of active inflammatory processes (e.g. trauma, infection), therefore subjects with hs-CRP>10 mg/l were excluded from the analysis (n=16). After exclusion for high hs-CRP levels and missing PWV and AIx data, the final study population comprised 362 men. We examined the normality assumptions for the variables by the Kolmogorov-Smirnov test. We used one-way ANOVA or Kruskal-Wallis test, according to the distributions of variables. Differences in proportions were tested using χ2 statistic. Linear regression analysis was used to assess the relationship between serum hs-CRP levels and PWV or AIx as the dependent variable. Quintiles of serum hs-CRP levels were used as dummy variables in the models because of its skewed distribution. Crude models were adjusted in three steps, first we added the risk factors which are always necessary for arterial stiffness in order to take into account the effect of age, mean arterial pressure, pulse pressure, and heart rate. Additionally, we adjusted for the classical cardiovascular risk factors waist-to-hip ratio, prevalent CVD, diabetes mellitus, hypertension, smoking status, total cholesterol, triglyceride, HDL and in the case of AIx also for height. To limit the effect of subclinical atherosclerosis on the probable association of hs-CRP and arterial stiffness, additional adjustments were made for CIMT. The AIx was finally adjusted for PWV. Furthermore, subgroup analysis was performed for anti-hypertensive and anti-lipid medication use (yes/no). The data were analyzed using the SPSS statistical package for Windows version 12.0 (SPSS Inc., Chicago, IL, USA). Results________________________________________________ _________________________ Table 1 shows baseline characteristics of the study population. Mean age of subjects was 60 years (range 40 to 80) and the median hs-CRP-level was 1.3 mg/l (interquintile range 0.60-2.70). The cutoff points for the quintiles of hs-CRP were 0.6, 1.1, 1.9, 3.6 mg/l. Table 2 shows the distribution of potential confounding variables across quintiles of hs-CRP. Subjects with the higher serum hs-CRP levels were older and had higher proportion of hypertension, diabetes mellitus, smoking, previous history of cardiovascular disease.

59

Chapter 3

Table 1. General characteristics of the study population. Variable Age (year) Systolic blood pressure (mmHg) Diastolic blood pressure (mmHg) Mean arterial blood pressure (mmHg) Pulse frequency (beats/min) Waist –to-hip ratio Total cholesterol (mmol/l) HDL (mmol/l) CIMT(mm) PWV (m/s) AIx (%) hs -CRP (mg/l) Triglycerides (mmol/l) Hypertension DM Current smoking CVD Blood pressure lowering medication Lipid lowering agent

60

N=362 Mean±SD 60±11 143.4 ±22.2 81.5±10.3 102.1±12.6 64±10 1.0±0.1 5.8±1.1 1.3±0.3 0.99±0.21 9.41±2.71 24.53±10.25 Median (interquartile range) 1.30(0.60-2.70) 1.30(0.95-1.80) N (%) 226(62.4) 38(10.5) 87(24.0) 48(13.3) 58(16.0) 45(12.4)

Hypertension DM Current smoking CVD Blood pressure lowering medication Lipid lowering medication

hs-CRP (mg/l) Triglyceride (mmol/l)

Age (year) Systolic blood pressure (mmHg) Diastolic blood pressure (mmHg) Mean arterial blood pressure (mmHg) Pulse frequency (beats/min) Waist-to-hip ratio Total cholesterol (mmol/l) HDL (mmol/l) CIMT(mm) PWV (m/s) AIx (%)

2nd (n=79) Mean±SD 61.2±11.3 143.7±22.4 81.2±11.0 102.0±13.1 63±10 0.97±0.04 5.81±1.17 1.35±0.36 1.01±0.20 9.36±3.30 24.60±8.32 Median (Interquartile Range) 0.80(0.70-0.90) 1.20(1.00-1.80) N (%) 50(63.3) 6(7.6) 18(22.8) 15(19.0) 19(24.1) 13(16.5)

1st (n=76) Mean±SD 53.7±11.2 137.4±20.8 78.1±8.7 97.8±11.0 62±10 0.94±0.06 5.70±0.92 1.40±0.35 0.91±0.18 8.08±1.74 19.96±11.98 Median (Interquartile Range) 0.40(0.20-0.50) 1.00(0.80-1.48) N (%) 38(50) 2(2.6) 15(19.7) 3(3.9) 6(7.9) 7(9.2)

Mean±SD 59.9±11.4 145.2±17.9 83.5±10.9 104.0±11.8 63±10 0.98±0.06 5.88±0.99 1.30±0.30 0.99±0.20 9.55±2.43 25.42±10.27 Median (Interquartile Range) 1.40(1.20-1.70) 1.30(1.00-1.70) N (%) 55(71.4) 6(7.8) 14(18.2) 5(6.5) 12(15.6) 6(7.8)

3rd (n=77)

hs-CRP quintiles

Table 2. Characteristics of the study population according to quintiles of hs-CRP (mg/l).

4th (n=73) Mean±SD 63.5±8.9 143.6±21.5 82.1±8.9 102.6±11.2 67±11 0.99±0.05 5.94±1.02 1.30±0.39 1.05±0.21 9.83±2.52 26.26±9.43 Median (Interquartile Range) 2.60(2.10-3.0) 1.40(1.00-1.95) N (%) 49(67.1) 11(15.1) 13(17.8) 10(13.7) 13(17.8) 7(9.6)

5th (n=57) Mean±SD 62.2±11.2 146.9±27.4 82.9±11.2 104.6±15.4 65±11 0.99±0.06 5.79±1.34 1.21±0.25 1.01±0.23 10.56±2.86 27.13±9.59 Median (Interquartile Range) 5.80(4.50-7.45) 1.40(1.00-2.15) N (%) 34(59.6) 13(22.8) 27(47.4) 15(26.3) 8(14.0) 12(21.1)

3 times the upper limit of reference value) or renal diseases (creatinine > 180 μmol/l) or hematological abnormalities (hemoglobin ≤ 7.0 mmol/l, hematocrit ≥ 0.50) (n = 27 exclusions). There were still 47 participants who met one of the other exclusion criteria and 25 participants who did not want to start intervention for other reasons. If participants did not match the study-profile due to one or more exclusion criteria, appropriate steps were undertaken to refer (if necessary) the participant to his general practitioner (n = 96). Major reasons for exclusion were testosterone level > cut-of-point, fasting glucose > 6.9 mmol/l or an elevated PSA level. Also, a fasting blood sample for a specific panel of laboratory assessments (including a spare DNA blood sample) was taken. Finally, 241 men proceeded to the randomization visit. Randomization visit The randomization visit assessment was conducted when all lab results were known and always within 4 weeks from the information and screening visit. First, digital rectal examination and transrectal ultrasound of prostate were performed. If there was any suspicion regarding prostate pathology, appropriate steps were undertaken to refer the participant to its general practitioner or consulting urologist. Four men were excluded for this reason. If the rectal ultrasound did not show any signs of pathology the participant continued the following baseline measurements: medical history, family medical history, vital signs (blood pressure, pulse), physical examination, anthropometry measurements (height and body weight, waist- and hip circumference, upper leg-, arm- and calf circumference, sagittal abdominal diameter), International Prostate Symptom Score (IPSS) questionnaire, functional mobility measurements, bone mineral density measurements via DEXA scan, health related quality of life questionnaires and utilities instrument, measuring cognition (15-words test, Digit symbol Substitution test, Concept Shifting Task test, the Benton Judgment of Line Orientation test and the Shephard rotation task), full body DEXA scan (lean body mass, fat free and fat mass), Pulse Wave Velocity (PWV), abdominal ultrasound for fat distribution. Finally 237 men were eligible for randomization. These subjects were randomly assigned to the intervention or the placebo group. A randomization list was computer-generated by Organon N.V, Oss, The Netherlands. One box with active medication and one box with placebo medication were delivered at the UMC Utrecht Pharmacy with the randomization list. Pharmacy personnel labeled the jars for the participants and provided the study medication upon prescription of the trial physician. Randomization numbers were assigned to the subjects in orders of enrolment into the trial.

91

Chapter 4 Intervention The intervention consisted of four capsules of 40 mg testosterone undecanoate (Andriol ® Testocaps®) provided by Organon NV, Oss, the Netherlands. The placebo was an identically looking and tasting capsule. The duration of the intervention was 26 weeks in which participants had to take the supplement on a daily basis. If for one reason unbinding should be necessary during the course of the trial, a backup hospital pharmacist was available who was informed about the trial but not involved. Control phone call (6 weeks) All participants were called at 6 weeks after randomization. At 6 weeks participants were asked about: Medical history update, including co-medication and adverse events. Also, they filled in the IPSS questionnaire. Control visit (13 weeks) During the intervention period of 26 weeks, subjects were asked to visit our clinic at 13 weeks after randomization (table 2). At 13 weeks the following measurement were done: Medical history update, including co-medication, adverse events, vital signs, (blood pressure, pulse), digital rectal examination, IPSS questionnaire and laboratory measurements (PSA, hematology, liver and renal functions, spare blood sampling for additional investigations). Final visit (26 weeks) The final visit took place after 26 weeks of intervention. At the final visit, all tests carried out at baseline were repeated following the same procedures. End point measurements An overview of visits and measurements is shown in Tables 2 and 3, respectively. Endpoints were assessed at baseline and after 26 weeks. All the assessments took place at the baseline randomization visit were repeated at 26 weeks.

92

Trial design and baseline characteristics Table 2. Baseline characteristics of participants.

Age (yr) Weight (Kg) Body Mass Index (kg/m²) Smokers Alcohol users Hypertension Systolic Blood Pressure (mm/Hg) Diastolic Blood Pressure (mm/Hg) Pulse Pressure (mm/Hg) Pulse Wave Velocity (m/s) Total Testosterone (nmol/l) Free Testosterone (nmol/l) Bioavailable Testosterone (nmol/l) SHBG (nmol/l) Albumin (g/l) Cholestrol (mmol/l) HDL (mmol/l) LDL (mmol/l) Insuline (mIU/l) C-reactive proteine (mg/l) PSA (µ g/l) Creatinine (µ mol/l) ASAT (U/l) ALAT (U/l) AF (U/l) GGT (U/l) Hemoglobine (mmol/l) Hematocrit (%) Prostate Volume (ultra sound) (cc) IPSS

Testosterone (n=120) 67.3 ± 5.1 86.1 ± 13.3 27.54 ± 3.85 21 (17.5) 99 (82.5) 75 (62.5) 156.2 ± 23.2 89.8 ± 12.0 66.4 ± 15.9 10.00 ± 2.51 10.93 ± 2.06 0.22 ± 0.05 5.23 ± 1.15 33.17 ± 10.59 43.94 ± 2.31 5.61 ± 0.99 1.16 ± 0.28 3.92 ± 0.91 10.14 ± 9.50 4.27 ± 6.56 1.54 ± 1.1 93.35 ± 18.0 22.98 ±8.1 26.36 ± 11.0 71.57 ± 19.2 29.51 ± 15.751 9.18 ± 0.5 0.45 ± 0.0 28.20 ± 12.4 6.38 ± 5.00

Placebo (n=117) 67.5 ± 5.0 84.4 ± 13.6 27.20 ± 3.90 15 (12.8) 85 (77.3) 69 (59) 151.1 ± 22.6 86.8 ± 11.7 64.3 ±15.2 9.53 ± 2.66 10.50 ± 1.89 0.21 ± 0.05 5.04 ± 1.20 32.90 ± 10.38 43.80 ± 2.38 5.50 ± 0.97 1.16 ± 0.29 3.80 ± 0.87 8.73 ±5.41 4.09 ± 6.97 1.63 ± 1.1 93.72 ± 15.2 24.21 ± 12.2 26.74 ± 13.6 69.91 ± 17.9 30.18 ± 19.9 9.14 ± 0.6 0.45 ± 0.0 27.6 ± 9.8 6.50 ± 4.80

Values are mean± SD or number (percentages).

93

94 X X

Functional mobility

Cognition tests

Quality of Life (QoL)

X X X

Sexual functioning questionnaire

IPSS questionnaire

Randomization

ADAM/AMS questionnaire

X X

Pulse wave velocity (PWV)

X X

Ultrasound abdominal fat mass

DEXA full body

DEXA bone mineral density

X X

Rectal Toucher

X

X

Spare blood X

X

Hematology

Rectal Ultrasound

X

X

PSA X

X

Testosterone

Renal functions

X

DNA blood sample

Liver functions

X X

Glucose

X

Anthropometry

Blood sample

X

Vital signs

X

Screening and randomization visit V2 X

X

X

Information visit and inclusion criteria V1 X X

Physical examination

Medical history (/update)

Check in & exclusion criteria Informed consent

Assessment

Table 3. Overview of the visits and measurements.

X

X

Phone call 6 weeks

X

X

X X X X X

X

X

X X X X X X X X X X X X X X X X X X

X X X X X X

V4

V3

X

Final visit 26 weeks

Visit 13 weeks

Chapter 4

Trial design and baseline characteristics Functional mobility measurement We assessed the functional mobility by the use of timed "Get Up and Go" test and a questionnaire for the ability to perform activities of daily life the Stanford Health Assessment Questionnaire (HAQ) 25, 26. Furthermore, skeletal muscle strength was assessed measuring handgrip strength and isometric knee strength 27,28. During the timed "Get Up and Go" test, the time taken by an individual to rise from a standard chair, walk three meters, turn around, return and sit down again was measured. The subject was requested to sit with his back against the chair and arms resting on the chair and performs the test three times. The fastest time was recorded in seconds. The Stanford Health Assessment Questionnaire (HAQ) has been widely used to measure functional status and includes 24 questions grouped into 8 categories of 2, 3 or 4 ADL's. The categories were dressing, arising, eating, walking, hygiene, reaching, gripping, and others. Participants responded to these questions by checking the level of difficulty from 0 (without any difficulty) till 3 (unable to do). If participants needed help from another person or assistive devices for each of the ADL's the score raises automatically to 2 (with much difficulty). Handgrip strength was measured with the JAMAR® dynamometer. The size of the grip was set so that the participant felt comfortable. The participant was in standing position his shoulder was adducted and neutrally rotated; the arm was vertical and the wrist in a neutral position. The participant squeezed the grip with maximal strength, alternating the left and right hand. The unit was automatically recorded the highest force exerted. Each test was repeated at least 5 times until no further improvements were seen. The best measure, recorded in kilograms, was used for analysis. Isometric knee extensor strength was measured with a hand-held dynamometer. The participant was in a seated position at a mat-table with the hip flexed to 90 degrees, the knee stretched to 180 degrees and the legs dependent. The dynamometer was applied perpendicularly to each lower extremity just proximal to the malleoli. Participants were instructed to take a second or two to come to maximum effort and to then push as hard as possible during another three seconds, while the investigator was giving counterforce. Each test was repeated five times, and if the examiner was not confident that a maximal effort was reached one more effort was made. The best measure, recorded in Newton, was used for analysis. Quality of life and well-being measurement Quality of Life and well being was measured by the Short Form-36 Health Survey as a generic QoL questionnaire (SF-36) and the Herschbach-questionnaire as a hormone specific questionnaire. The SF-36 is a questionnaire consisting of questions regarding general health, ability to perform physical activity and work, emotional problems and assessment of his own health 29. The Herschbach questionnaire is a questionnaire translated from the questionnaire "Fragen zur Lebenszufriedenheit" (FLZ) according to the method described by Henrich and Herschbach. The questionnaire is divided in a "general" and a "health" section, each including eight items. All items have been evaluated on a 5-

95

Chapter 4 point scale according to their individual importance (I) and degree of satisfaction (S). As a measure of evaluation, a combination of importance and satisfaction (I-1)*(S*2–5) will be used. In addition the sum of the combination values will be calculated for each section 30. Sexual behavior and erectile dysfunction measurement The 'Eleven questions about sexual functioning (ESF) questionnaire, developed by the National Institute for Social Sexual Research (Rutgers Nisso Group, Utrecht, The Netherlands), has been used to assess sexual well being. The questionnaire has 11 questions measuring sexual drive (two questions); erectile function (three) and ejaculatory function (two), as well as assessing problems with sex drive, erections, or ejaculation (three); and overall satisfaction with sex life (one). Each question is scored on a scale of 0–4, with higher scores indicating better functioning. The Androgen Deficiency in Ageing Males (ADAM) and the Ageing Males' Symptoms rating scale (AMS) questionnaires have been administered as well. The ADAM questionnaire contains 10 questions regarding the age-related decline in androgens. All questions should be answered yes or no. A positive questionnaire result, indicating an androgen deficient state, is defined as a 'yes' answer to question 1 or 7 or any 3 other questions 31. AMS-questionnaire (Ageing Males' Symptoms rating scale) is a 17-question questionnaire investigating age related health complaints divided in three dimensions (psychological, somatovegetative and sexual) of each 5 questions. Each question can be scored from 1 (no symptoms) to 5 (very severe symptoms), so a dimension can score from 5 to 25 points. Within each dimension, cumulative scores indicate the severity of the complaints on each territory; also the cumulative of all dimensions indicate an overall view of Ageing Males' Symptoms. Classification range spreads from no impairment at all to severe impairment 32. Body composition measurement Body Composition was assessed by anthropometry (body mass index (BMI), waist and hip girth, upper arm-, upper leg and calf circumference and sagittal abdominal diameter), full body DEXA scan (lean body mass, fat free- and fat mass) and ultrasound of the abdominal fat mass. BMI was calculated as the weight in kilograms divided by the square of the height in meters, after taking of coat, sweaters and shoes. All circumference measurements were done with a standard household centimeter. Waist circumference was measured at the level of midway the distance between the lower rib and iliac crest, after normal expiration without pressure of the centimeter at the skin. The hip circumference was measured at the level of the greater trochanter. The upper arm circumference was measured at the non-dominant arm at the level of midway between the tip of the acromnion and the olecranon. The thigh circumference was measured just below the gluteal fold of the left leg. Calf circumference was measured at the level of the largest circumference of the left calf. The Sagittal abdominal diameter (SAD) was measured using a HoltainKahn (abdominal calliper (Holtain ltd., Crosswell, UK) which allows a direct reading of the distance between the subjects back and the front of the subjects' abdomen. With subject in supine position a

96

Trial design and baseline characteristics mark was made halfway between the left and right iliac crest. The lower arm of the caliper was inserted underneath the subjects back and the upper arm was adjusted until touching the abdominal wall at the level of the mid-abdominal mark. The measurement was taken with a resting and at the end of a normal expiration. The distance between the subjects back and abdominal wall was measured on a centimeter scale and round off to the nearest 0.1 cm. Since with the abdominal subcutaneous fat tends to slip along the flanks, when the subject is in supine position, the SAD is an indirect measurement of the amount of visceral fat mass. Total body composition was measured with dual energy X-ray absorptiometry (DEXA) using a Lunar prodigy® DEXA instrument. Scanning was performed according to the instructions of the manufacturer. After placement of the subject on the table, there was scanning of the whole subject from dorsal to ventral. Both legs and feet were endorotated and fixed to on another. Calculations were made regarding fat-mass, fat-free mass and lean body mass 33. Abdominal ultrasonography was performed in all abdominal obese subjects with a Ultramark 9®. The distances between the posterior edge of the abdominal muscles and the lumbar spine or psoas muscles were measured using electronical callipers. For all images the transducer was placed on a straight line drawn between the left and right mid-point of lower rib and iliac crest. The middle was marked 10 cm from the left and right side. Distances were measured from three different angles: medial, left and right for intra-abdominal fat mass and medial for subcutaneous fat mass. Measurements were made at the end of quiet expiration, applying minimal pressure without displacement of intra-abdominal contents as observed by ultrasound image 34. Cognitive function measurement Cognitive function was measured as follows. Verbal memory was tested with the Dutch version of the Rey Auditory Verbal Learning Task. This is a test for long-term memory retention. Fifteen words were read to the subject, who was required to report as many words as he could remember immediately after presentation. After a delay of 15 min (in which another test, the Benton Judgment of Line Orientation, was administered), the subject was asked to recall as many words as possible from memory 35. Mental processing speed was tested with the "Digit Symbol Substitution test". This is a subtest from the Wechsler Adult Intelligence Scale (WAIS) that covers general knowledge. It measures cognitive and perceptual-motor processing speed. The subject was given a code that pairs symbols with digits. The test consists of matching as many series of digits to their corresponding symbols as possible in 90 sec 36. The trail making test was used to test planning of movement, vasomotor tracking, and processing speed. In this test, pseudo-randomly placed circles with numbers (Trail Making A1), with letters (Trail Making A2), and with both numbers and letters (Trail Making B) have to be connected with a line as fast as possible in a fixed order. In the event of error, the subjects were immediately informed and asked to restart from the point of error: this was done with the timer left running. The time taken to complete the trail without error was recorded 37. The "Benton Judgment of Line

97

Chapter 4 Orientation test" was used to measure visual-spatial skills. This test measures basic perceptual processes contributing to extra-personal spatial perception. The test requires the subject to identify which 2 of 11 lines presented in a semicircular array have the same orientation in two-dimensional space as two-target lines 38. The visospatial performance was assessed by the Vandenberg and Kuse adaptation of Shepard and Metzler's three-dimensional mental rotations test 39. This is the cognitive task that has been most consistently associated with testosterone levels. The test was consisted of 20 items in which the subject was presented with a three-dimensional geometric target line drawing and four test drawings, and was required to indicate which two of four test drawings depict the target drawing in rotated positions. Two parallel test versions were made by taking the odd and even items on time 1 (baseline) and time 2 (after intervention) respectively (10 items for each test). These parallel versions have been shown to correlate strongly with each other and to have a high reliability. Subjects were instructed to "work as quickly as possible, but do not sacrifice accuracy for speed". They were allowed 10 minutes to complete the test. Aortic stiffness and cardiovascular risk factors measurements Total cholesterol, HDL cholesterol and triglyceride were measured by a timed endpoint method (Synchron LX®, Beckman Coulter, Fullteron, California, USA) 40. LDL was calculated with the Friedewald equation 41. Insulin was measured by a solid-phase two site chemiliminesent immunometric assay (IMMULITE 2000, Diagnostic Products Corporation, Los Angeles, California, USA). Serum levels of highly sensitive CRP were measured using a near-infrared particle immunoassay of the Synchron LX System (Synchron LX®, Beckman Coulter, Fullteron, California, USA). Systolic and diastolic blood pressures and pulse were measured in duplicate at the dominant arm with the subjects in sitting position after 5 minutes of rest with an automated and calibrated oscillometric device (Omron Healthcare Europe, Hoofddorp, The Netherlands). Subsequently, the mean systolic and diastolic blood pressures and mean pulse rate were calculated. Aortic stiffness was determined by means of pulse wave velocity. The Sphygmocor® system was used to non-invasively measure stiffness of the aorta (Pulse wave velocity system, PWV medical, Sydney, Australia) 42. After 5 to 10 minutes rest of the subject in supine position, aortic PWV was measured by sequentially recordings of arterial pressure waveform at the carotid artery and the femoral artery using a hand-held micromanometer-tipped probe on the skin at the site of maximum arterial pulsation. Gating the recordings at those two sites to the electrocardiogram (ECG) allowed PWV to be measured. Recordings were taken when a reproducible signal was obtained with high amplitude excursion, i.e. usually 10 consecutive beats to cover complete respiratory cycle. The system software, using the R wave of a simultaneously recorded ECG as a reference frame, was calculated the wave transit time. A distance from the carotid-sampling site to the suprasternal notch and suprasternal notch to the femoral artery was measured using a compass 43.

98

Trial design and baseline characteristics Bone mineral density measurement Bone mineral density (BMD) was measured with dual energy X-ray absorptiometry (DEXA) using a Lunar prodigy® DEXA instrument. Scanning was performed according to the instructions of the manufacturer. BMD was measured of lumbar vertebrae (L1–L4 individually and together) and proximal femur (femoral neck, trochanter, inter-trochanter, Ward's triangle and total hip, left-or right if left not available). A T-score≤-2.5 denotes osteoporosis, a T-score between -1 and -2.5 denotes osteopenia 44. The DEXA scan was also used to measure total and trunk lean body mass (see total body DEXA-scan). Quality assurance, including calibration was performed routinely every morning for DEXA (if that day a measurement is planned), using the standard provided by the manufacturer 45. Prostatic measurements Effects on the treatments on the prostate were examined by digital rectal examination, transrectal ultrasound of the prostate and by monitoring serum prostate-specific antigen (PSA) levels and by IPSS. The IPSS, developed by the American Urological Association (AUA), contains seven items that measure frequency and severity of urological symptoms, together with an additional item measuring the overall impact of these symptoms on quality of life. Each of the seven symptom items has a response scale with six choices, scored from 0 (absence of the symptom) to 5 (symptom always present). Symptoms are considered mild for scores between 0 and 7, moderate for scores between 8 and 19, and severe for scores between 20 and 35 46. Digital rectal examination was performed at baseline, 13 weeks and at the end of treatment (26 weeks). Biplanal transrectal ultrasonography of the prostate was performed at baseline and at the end of treatment (26 weeks) with a 7-MHZ transrectal probe (Bruel and Kjaer Model 2110 Falcon). If rectal ultrasound was abnormal, patients were excluded and referred for further evaluation. Serum prostate specific antigen (PSA) levels were measured by an immunnometric assay (IMMULITE® 2000 PSA, Diagnostic Products Corporation, Los Angeles, California, USA) at baseline, week 13 and at the end of the study. An increase of > 1.4 μg/l between measurements at any time was cause for concern. Abnormal values required repeat testing; if values remained high, co-morbid illness was ruled out. This was reason to exclude a patient and send to his general practitioner. Laboratory measurements Fasting blood samples were obtained between 8.00 and 11.00 AM to minimize diurnal variation. The level of total testosterone and sex hormone binding globulin (SHBG) were measured with a solidphase, competitive, chemiluminescent enzyme immunoassay (IMMULITE® 2000, Diagnostic Products Corporation, Los Angeles, California, USA) at baseline and at the end of the study. The intra-assay coefficient of variation of this assay was 7.2% and the inter-assay coefficient of variation was 8.2 % for testosterone and 2.5% and 5.2 % for SHBG, respectively. Hematology (hemoglobin and hematocrit) and routine biochemistry (liver functions and creatinine) were measured by standard

99

Chapter 4 autoanalyzer methodologies (Synchron LX®, Beckman Coulter, Fullteron, California, USA) at baseline, after 13 weeks, and at the end of the study. During the study, hemoglobin levels of ≤ 7 mmol/l, hematocrit levels ≥ 0.50, liver function values ≥ three times normal upper normal reference level (ASAT: 15–45 U/l; ALAT: 10–50 U/l; AF: 40–130 U/l ; γ-GT: 15–70 U/l), or creatinine levels of ≥ 180 μmol/l led to an extra blood check after a week. If the values were still too high study participation was discontinued. All laboratory measurements were done at the SHO laboratory, Velp, The Netherlands. Adverse events An adverse event (AE) was defined as any untoward medical occurrence in a patient or clinical investigation subject administered a pharmaceutical product and which did not necessarily have a causal relationship with this treatment. An AE could therefore be any unfavorable and unintended sign (including an abnormal laboratory finding), symptom, or disease temporarily associated with the use of a medicinal report, whether or not considered related to the medicinal product. Whether or not an abnormal laboratory/vital sign were entered on the AE form depends on whether or not the finding was clinically relevant in the opinion of the investigator. Information regarding AEs was obtained by questioning or examining the subject. At each visit during the treatment period, all new complaints and symptoms (i.e. those not existing before the treatment period) were recorded (and coded) on the AE Form. Pre-existing complaints or symptoms that increased in intensity or frequency during the treatment period were entered on the AE Form also. All AEs were characterized in terms of their start and stop dates, maximum intensity, action taken on trial medication, relationship to trial medication, and subject outcome. If a subject discontinued the trial because of an AE, this was noted on the AE Form. Serious adverse events (SAE) forms were supplied by Organon. If the AE meets the definition of an SAE, the procedure for reporting SAEs was followed. A serious adverse event (SAE) was defined as any untoward medical occurrence that at any dose: resulted in death, was life-threatening, required in-patient hospitalization or prolongation of existing hospitalization, resulted in persistent or significant disability or incapacity, noted the term life threatening refers to an event in which the patient was at risk of death at the time of the event; it did not refer to an event, which hypothetically might have caused death if it had been more severe. Medical and scientific judgment was exercised in deciding whether expedited reporting was appropriate in other situations, such as important medical events that might not be immediately life-threatening or result in death or hospitalization, but may jeopardized the patient or might required intervention to prevent one of the other outcomes listed in the definition above. These were considered serious. All SAEs were reported to the METC and to Organon NV, Oss, the Netherlands. Every attempt was made to obtain any relevant laboratory or hospital reports that pertain to the SAE.

100

Trial design and baseline characteristics Compliance Compliance was monitored by spare capsule counting at each study visit. After finalization of the study serum testosterone concentrations were assessed in the final visit blood samples as an extra check on compliance. Power calculation The pre-specified number of subjects was 240 in total, 120 in each intervention arm. This number was based on conventional assumptions of α = 0.05 and β = 0.20, withdrawal from intervention of 15% and an improvement of 25% on MHAQ and of 18% on the 15 Words test. These improvements were realistic, since they have been previously reported in short-term small studies. Data analysis The primary analysis will be done by linear regression analysis with change in outcome parameter between final visit and baseline visit as the dependent and treatment group as the independent variable. All analyses will be based on an intention-to-treat approach (i.e., the intention-to-treat group will consist of all subjects, including those who withdrew from blinded medication, who received at least one dose of study drug and who had at least one post-baseline assessment of the outcome variable). In addition to an intention-to-treat analysis, a per-protocol analysis will be performed. The per-protocol group will consist of all subjects from the intention-to-treat group who did not have any major protocol violations. Furthermore, subgroup analysis will be performed for the following predefined subgroups according to baseline measurements: waist girth (< 100 cm versus ≥ 100 cm); testosterone level (< 12 versus ≥ 12 nmol/l), age (< median versus ≥ median), and baseline level of outcome under study (< median versus ≥ median). Differences between final visit and baseline for continuous measures were expressed as means and 95% confidence intervals; unpaired t-tests were used for testing. Level of significance was set at P < 0.05. All analyses are performed with SPSS, statistical software package, version 11. References____________________________________

___ _____________________________

1.

Lamberts SWJ, van den Beld AW, van der Lely A-J. The Endocrinology of aging. Science 1997; 278:419-424.

2.

Harman SM, Metter EJ, Tobin JD, Pearson J, Blackman MR. Longitudinal effects of aging on serum total and free testosterone levels in healthy men. J Clin Endocrinol Metab 2001; 86:724-731.

3.

Vermeulen A. Andropause. Maturitas 2000, 34:5-15.

4.

Greendale G, Edelstein S, Barrett-Connor E. Endogenous sex steroids and bone mineral density in older women and men: the Rancho Bernardo Study. J Bone Miner Res 1997; 12:1833-1843.

101

Chapter 4 5.

Matsumoto AM: Andropause: clinical implications of the decline in serum testosterone levels with aging in men. J Gerontol A Biol Sci Med Sci 2002; 57:76-99.

6.

Ferrando AA, Sheffield-Moore M, Paddon-Jones D, Wolfe RR, Urban RJ. Differential anabolic effects of testosterone and amino acid feeding in older men. J Clin Endocrinol Metab 2003; 88:358-362.

7.

Muller M, Aleman A, Grobbee DE, de Haan EHF, van der Schouw YT. Endogenous sex hormone levels and cognitive function in aging men: Is there an optimal level? Neurology 2005; 64:866-871.

8.

Snyder PJ, Peachey H, Hannoush P et al. Effect of testosterone treatment on bone mineral density in men over 65 years of age. J Clin Endocrinol Metab 1999; 84:1966-1972.

9.

Muller M, van der Schouw YT, Thijssen JHH, Grobbee DE. Endogenous sex hormones and cardiovascular disease in men. J Clin Endocrinol Metab 2003; 88:5076-5086.

10.

Burris A, Banks S, Carter C, Davidson J, Sherins R. A long-term, prospective study of the physiologic and behavioral effects of hormone replacement in untreated hypogonadal men. J Androl 1992; 13:297-304.

11.

Snyder PJ, Peachey H, Berlin JA et al. Effects of testosterone replacement in hypogonadal men. J Clin Endocrinol Metab 2000; 85:2670-2677.

12.

Couillard C, Gagnon J, Bergeron J et al. Contribution of body fatness and adipose tissue distribution to the age variation in plasma steroid hormone concentrations in men: The HERITAGE Family Study. J Clin Endocrinol Metab 2000; 85:1026-1031.

13.

Breuer B, Trungold S, Martucci C et al. Relationships of sex hormone levels to dependence in activities of daily living in the frail elderly. Maturitas 2001; 39:147-159.

14.

Liverman C, Blazer D, Editors. Testosterone and aging: clinical research directions. Washington DC: The National Academies Press; 2004.

15.

Harman S, Naftolin F, Brinton E, Judelson D. Is the estrogen controversy over? deconstructing the Women's Health Initiative study: a critical evaluation of the evidence. Ann NY Acad Sci 2005; 1052:43-56.

16.

Kaufman JM. The effect of androgen supplementation therapy on the prostate. Aging Male 2003; 6:166-174.

17.

Hafez B, Hafez E. Andropause: endocrinology, erectile dysfunction, and prostate pathophysiology. Arch Androl 2004; 50:45-68.

18.

Gruntmanis U, Braunstein G. Treatment of gynecomastia. Curr Opin Investig Drugs; 2:643-649.

19.

Lanfranco F, Gianotti L, Maccario M. Endocrine and metabolic alterations in obstructive sleep apnea syndrome. J Endocrinol Invest 2003; 26:491-492.

20.

Dickerman R, Pertusi R, Zachariah N, Schaller F. Androgen-induced erythrocytosis. Am J Hematol 1998; 59:263264.

102

Trial design and baseline characteristics 21.

Shaneyfelt T, Husein R, Bubley G, Mantzoros CS. Hormonal predictors of prostate cancer: a meta-analysis. J Clin Oncol 2000; 18:847-853.

22.

Eaton N, Reeves G, Appleby P, Key T. Endogenous sex hormones and prostate cancer: a quantitative review of prospective studies. Br J Cancer 1999; 80:930-934.

23.

Schroder FH. The prostate and androgens: the risks of supplementation. In androgens and the aging male. Edited by Oddeens BJ VA. New York: The Partenon Publisher Group; 1996:223-232

24.

Gooren L. A ten-year safety study of the oral androgen testosterone undecanoate. J Androl 1994; 15:212-215.

25.

Pincus T, Summey J, Soraci SJ, Wallston K, Hummon N. Assessment of patient satisfaction in activities of daily living using a modified Stanford Health Assessment Questionnaire. Arthritis Rheum 1983; 26:1346-1353.

26.

Podsiadlo D, Richardson S. The timed "Up & Go": a test of basic functional mobility for frail elderly persons. J Am Geriatr Soc 1991; 39:142-148.

27.

Hamilton A, Balnave R, Adams R. Grip strength testing reliability. J Hand Ther 1994; 7:163-170.

28.

Samson M, Meeuwsen I, Crowe A, Dessens J, Duursma S, Verhaar HJJ. Relationships between physical performance measures, age, height and body weight in healthy adults. Age Ageing 2000; 29:235-242.

29.

Ware JJ, Sherbourne C. The MOS 36-item short-form health survey (SF-36). I. Conceptual framework and item selection. Med Care 1992; 30:473-483.

30.

Huber D, Henrich G, Herschbach P. Measuring the quality of life: a comparison between physically and mentally chronically ill patients and healthy persons. Pharmacopsychiatry 1988; 21:453-455.

31.

Morley JE, Charlton E, Patrick P et al. Validation of a screening questionnaire for androgen deficiency in aging males. Metabolism 2000; 49:1239-1242.

32.

Heinemann L, Zimmermann T, Vermeulen A, Thiel C, Hummel W. A new aging males symptoms rating scale. The Aging Male 1999; 2:105-114.

33.

Van Loan M, Mayclin P. Body composition assessment: dual-energy X-ray absorptiometry (DEXA) compared to reference methods. Eur J Clin Nutr 1992; 46:125-130.

34.

Tornaghi G, Raiteri R, Pozzato C et al. Anthropometric or ultrasonic measurements in assessment of visceral fat? A comparative study. Int J Obes Relat Metab Disord 1994; 18:771-775.

35.

Lezak MD: Neuropsychological assessment. New York: Oxford University Press; 1995.

36.

Wechsler D: Manual for Wechsler Adult Intelligence Sclae. New York: The psychological corporation; 1955.

37.

Reitan RM, Wolfson D. The Halstead-Reitan neuropsychological test battery. In. Tucson, Arizona: Neuropsychology Press 1985

103

Chapter 4 38.

Benton A, Hamsher K, Varney N, Spreen O. Contribution to neuropsychological assessment. In. New York: Oxford University Press; 1983

39.

Vandenberg S, Kuse A. Mental rotations, a group test of three-dimensional spatial visualization. Percept Mot Skills 1978; 47:599-604.

40.

Bucolo G, David H. Quantitative determination of serum triglycerides by the use of enzymes. Clin Chem 1973; 19:476-482.

41.

Friedewald W, Levy R, Fredrickson D. Estimation of the concentration of low-density lipoprotein cholesterol in plasma, without use of the preparative ultracentrifuge. Clin Chem 1972; 18:499-502.

42.

Lehmann E, Hopkins K, Gosling R. Assessment of arterial distensibility by automatic pulse wave velocity measurement. Hypertension 1996; 27:1188-1191.

43.

O'Rourke M, Gallagher D. Pulse wave analysis. J Hypertens Suppl 1996; 14:147-157.

44.

Kanis JA. Diagnosis of osteoporosis and assessment of fracture risk. The Lancet 2002; 359:1929-1936.

45.

Hansen M, Hassager C, Overgaard K, Marslew U, Riis B, Christiansen C. Dual-energy x-ray absorptiometry: a precise method of measuring bone mineral density in the lumbar spine. J Nucl Med 1990; 31:1156-1162.

46.

Barry M, Fowler FJ, O’Leary M et al. The American Urological Association symptom index for benign prostatic hyperplasia. J Urol 1992;148:1549–1557.

104

Chapter 4.2. Effects of testosterone supplementation on blood pressure and aortic stiffness

Manuscript based on this chapter: Nakhai Pour HR, Grobbee DE, Emmelot-Vonk MH, Michiel L Bots, Verhaar HJJ, van der Schouw YT. Effects of a six-month oral testosterone supplementation on blood pressure and aortic stiffness in elderly men. (Submitted)

105

Chapter 4 Summary_______________________________________________________ _________________ Background: The cardiovascular effects of testosterone supplementation remains controversial and long-term data in larger samples are scarce. We set out to determine the impact of 6-months oral testosterone supplementation (Andriol®Testocaps®) on systolic blood pressure (SBP), diastolic blood pressure (DBP), pulse pressure (PP) and aortic stiffness in elderly men with moderately low testosterone levels. Subjects and methods: 237 community-dwelling healthy men aged 60-80 years with testosterone levels < 13.7 nmol/l were recruited into a double-blind randomized placebo-controlled trial and were randomized to either four capsules of 40 mg testosterone undecanoate (Andriol®Testocaps®) or placebo daily for six months. SBP and DBP were obtained using oscillometric device at baseline, 3 months, and final visit. Aortic stiffness was measured as carotid-femoral pulse wave velocity (PWV) at baseline and final visit. Results: After 6 months of therapy, compared to placebo, there were no significant improvements in SBP (1.9 mmHg CI -2.9; 6.8), DBP (1.1 mmHg CI -1.7; 3.8) and PP (0.9 mmHg CI -2.7; 4.4) in the testosterone group. Moreover, compared to placebo, 6 months of testosterone supplementation did not change the aortic PWV (0.3 m/s CI -0.3; 0.9). Conclusion: Testosterone supplementation, in dosage of 160 mg daily for six months, has no effect on blood pressure measurements and aortic stiffness in elderly men with a low normal testosterone concentration.

106

Testosterone supplementation and arterial blood pressure and stiffness Introduction______________________________________________________________________ The process of aging in men is accompanied by a decline in serum testosterone levels 1. This decline in testosterone levels by aging coincides with an increase of blood pressure and acceleration of large artery stiffness 2-9. Increased blood pressure and arterial stiffening are strong risk factors for cardiovascular morbidity and mortality, independently of age 10-12. Relatively little is known about the impact of decline in sex hormones levels on blood pressure. There are reports of lower circulating testosterone levels in men with hypertension 3,13. On the other hand, supplementation of testosterone has been related to hypertension in other studies 14,15. Stiffness is a dynamic component of arterial performance and it is susceptible to modification both in the short- and long term by several factors that affect the structure and/or function of the arterial wall 16,17. Stiffening of the arterial tree increases the cardiac afterload, and, by lowering the diastolic pressure, potentially compromises coronary perfusion 18-20. The role of testosterone in large artery stiffening is still under debate 7,21. Arterial stiffness is thought to be applicable as a marker relating with the severity of atherosclerosis and/or predicting future atherosclerotic cardiovascular events, therefore, we prospectively assessed the effects of a 6-month of oral testosterone supplementation on blood pressure measurements and aortic stiffness. Methods_____________________________________________________ ___________________ Design We conducted a randomized double-blind placebo-controlled trial. Details of the study design, recruitment, and procedures have been published previously 22. The Institutional Review Board of the University Medical Centre Utrecht approved the study protocol and all participants provided written informed consent. Participants Participants were recruited by direct mailing to 8020 randomly selected men between 60 and 80 years of age whose addresses were obtained from the municipal register of the city of Utrecht, the Netherlands. Inclusion criteria included a testosterone level below the 50th percentile of the study population based testosterone distribution and an age between 60 and 80 years. The 50th percentile cut-off level of testosterone was determined to be 13.7 nmol/l after screening of 50 candidates. Exclusion criteria included: a history of myocardial infarction or cerebrovascular accident within the past 6 months, cardiac failure unless medically treated and not symptomatic, malignancy within the past 5 years except for non-melanoma skin cancer, any hormone dependent tumour in the history, serious liver- or renal diseases, hematological abnormalities, epilepsy or the use of anti-epileptic medication, migraine more than once a month, diabetes mellitus, a fasting glucose level ≥ 6.9 mmol/l,

107

Chapter 4 corticosteroid use, use of testosterone esters and alike substances within the past 60 days, prostate hypertrophy in medical history, and an age-related elevated prostate-specific antigen (PSA) levels. Following an initial telephone contact, 684 men were screened with medical history, laboratory testing and digital rectal examination. Finally, 237 men were eligible for entry into the study and agreed to participate. The flow of study subjects participants recruitment and enrolment is shown in figure 1. Figure 1: Participant flow diagram.

Invited by mail to participate (n=8020)

Excluded (n= 6174) Did not return answer form (n=6142) Refused to participate (n=32)

Send information leaflet (n=1846) (n= 69) Did nor reach in time (n=51) Did not continue phone call (n=18) Contacted by telephone (n=1777) (n= 1093) Meeting exclusion criteria (n=346) Refused to participate (n=747) Invited for screening visit (n=684)

Randomized (n=237)

Assigned to receive testosterone (n=120)

Assigned to receive placebo (n= 117)

16 withdrew (7 without and 9 with follow up) 2 refused to take medicines/no improvement 1 lost to follow up 4 gastro-intestinal complaints 2 increase in PSA 1 increase in liver functions 1 cardiac complaint 1 prostatic complaints 4 other reasons

113 included in primary analysis 104 included in per-protocol analysis

108

(n= 447) Urologic obstruction (n=20) Heart failure (n=17) Further medical exclusion (n=10) Fasting Glucose >6.9 mmol/l (n=35) Testosterone > cut –off-point (n=279) Elevated PSA (n=30) Other laboratory exclusion (n=27) Abnormality of the prostate (n=4) Did not want intervention (n=25)

14 withdrew (7 without and 7 with follow-up) 2 refused to take medicines/no improvement 1 gynaecomastia 3 gastro-intestinal complaints 4 increase of PSA 2 cardiac complaints 2 other reasons

110 included in primary analysis 103 included in per protocol analysis

Testosterone supplementation and arterial blood pressure and stiffness Intervention At entry into the study, subjects were randomly assigned to treatment with either two capsules of testosterone undecanoate (Andriol®Testocaps®) 40 mg twice a day during the meal, or matching placebo, for a total duration of 6 months. Capsules were taken prior to the onset of breakfast and dinner. Adherence was monitored by pill counting at each study visit. After finalization of the study serum testosterone concentrations were measured in the final visit blood samples as an extra check on compliance. Measurements Systolic and diastolic blood pressures were measured in duplicate at the left arm with the subjects in sitting position after 5 minutes of rest with an automated and calibrated oscillometric device (Omron Healthcare Europe, Hoofddorp, the Netherlands). Subsequently, the mean systolic and diastolic blood pressures were calculated. Pulse pressure was calculated as systolic blood pressure minus diastolic blood pressure. Blood pressure was measured at baseline, after 13 weeks, and at the final visit. Aortic stiffness was determined by means of pulse wave velocity. The Sphygmocor® system was used to non-invasively measure stiffness of the aorta (Pulse wave velocity system, PWV medical, Sydney, Australia) 23 as has been successfully used in other studies 24,25. After 5 to 10 minutes rest of the subject in supine position, aortic PWV was measured by sequentially recordings of arterial pressure waveform at the carotid artery and the femoral artery using a hand-held micromanometer-tipped probe on the skin at the site of maximum arterial pulsation. Gating the recordings at those two sites to the electrocardiogram (ECG) allowed PWV to be measured. Recordings were taken when a reproducible signal was obtained with high amplitude excursion, i.e. usually 10 consecutive beats to cover a complete respiratory cycle. The system software calculated the wave transit time, using the R wave of a simultaneously recorded ECG as a reference frame. The distance from the carotid-sampling site to the suprasternal notch and suprasternal notch to the femoral artery was measured using a compass. PWV was measured at baseline and at the final visit. Laboratory Fasting blood samples were obtained between 08.00h and 11.00h am to minimize diurnal variation. The levels of total testosterone and SHBG were measured with a solid-phase, competitive, chemiluminescent enzyme immunoassay (IMMULITE ® 2000, Diagnostic Products Corporation, Los Angeles, California, USA) at baseline and at the end of the study. The intra- and inter-assay coefficients of variation were 7.2% and 8.2% for testosterone, and 2.5% and 5.2% for SHBG, respectively. Free and bioavailable testosterone was calculated from testosterone, SHBG and albumin using the method described by Vermeulen et al 26.

109

Chapter 4 Power calculation The study was a priori designed to assess the effect of testosterone on cognitive decline, and thus powered for two primary end points: the 15 Words test for cognitive function and the Health Assessment Questionnaire (HAQ) for performance in ADL. The planned number of subjects was 240 in total, 120 in each intervention arm 22. With these numbers and using conventional assumptions of alpha=0,05 and beta=0.20, we were able to demonstrate differences between treatment groups of 8 mmHg (5.4%) in systolic blood pressure, 4 mmHg (4.8%) in diastolic blood pressure, and of 0.9 m/s (9.0 %) in PWV. Statistical analyses The achieved values of outcome parameters (blood pressure (systolic, diastolic), pulse pressure and pulse wave velocity)) at the final visit were expressed as means and 95% confidence intervals. The primary analysis was done according to the modified intention-to-treat principle for those who had at least two measurements including baseline, in the groups to which they were randomized. Linear regression analysis was used with the achieved values at the final visit in outcome parameters as the dependent variable and group allocation as the independent variable. Comparison between treatment groups for blood pressure measurements change at baseline, 3 and 6 months was made using repeated measures analysis of variance (ANOVA). In addition a per-protocol analysis was performed in the subjects who complied with the study protocol for six months. Furthermore, subgroup analysis performed for subgroups according to baseline testosterone level (the lowest tertile versus the two highest tertiles), age (< median versus ≥ median), and baseline level of SBP, DBP, PP and PWV (< median versus ≥ median). The level of significance was set at P10 mg/l can be taken as evidence of active inflammatory processes (e.g., trauma, infection), therefore subjects with hs-CRP>10 mg/l at baseline or at the end of the study were excluded from the analysis (n=37). The levels of total testosterone and SHBG were measured with a solid-phase, competitive, chemiluminescent enzyme immunoassay (IMMULITE ® 2000, Diagnostic Products Corporation, Los Angeles, California, USA) at baseline and at the end of the study. The intra- and inter-assay coefficients of variation were 7.2% and 8.2% for testosterone, and 2.5% and 5.2% for SHBG, respectively. Free and bioavailable testosterone was calculated from testosterone, SHBG and albumin using the method described by Vermeulen et al 15. Data analysis We used the t test to evaluate differences in means and the χ2 statistic to evaluate differences in proportions. As the distributions of hs-CRP were skewed, differences in medians between treatment groups in final visit were tested by using the Wilcoxon rank sum test. All analyses were based on a modified intention-to-treat approach (i.e., the intention-to-treat group consists of all subjects, including those who withdrew from blinded medication, who received at least one dose of study drug and who had at least one post-baseline assessment of the outcome variable). In addition a per-protocol analysis was performed in the subjects who complied with the study protocol for 26 weeks. Furthermore, pre-specified subgroup analysis were performed according to baseline testosterone level (the lowest tertile versus the two highest tertiles), age (< median versus ≥ median), and baseline level of CRP (< median versus ≥ median). Differences between the values of hs- CRP in final visit were expressed as means and 95% confidence intervals. Level of statistical significance was set at P