E n d o c r i n e

Effects on Lipoprotein Particles of Long-Term Dehydroepiandrosterone in Elderly Men and Women and Testosterone in Elderly Men
Manivannan Srinivasan,* Brian A. Irving,* Robert L. Frye, Peter O’Brien, Stacy J. Hartman, Joseph P. McConnell, and K. Sreekumaran Nair
Division of Endocrinology, Endocrine Research Unit (M.S., B.A.I., K.S.N.), Cardiovascular Laboratory Medicine (S.J.H., J.P.M.), Division of Biostatistics (P.O.), Division of Cardiovascular Diseases (R.L.F.), Mayo Clinic and Mayo Foundation, Rochester, Minnesota 55905 Context: Although age-related declines in dehydroepiandrosterone sulfate (DHEAS) and testosterone are associated with cardiovascular risk, it remains to be determined whether replacement of these hormones improves cardiovascular risk factors. Objective: This study sought to determine the effect of long-term replacement of dehydroepiandrosterone (DHEA) in elderly men and women and testosterone in elderly men on lipid and lipoprotein concentrations and particle sizes. Methods: A 2-yr randomized, placebo-controlled, double-blind study was conducted in 87 elderly men with low levels of DHEAS and bioavailable testosterone and 57 elderly women with low levels of DHEAS. Among elderly men, 29 received DHEA (75 mg/d), 27 received testosterone (5 mg/d), and 31 received placebo. Among the elderly women, 27 received DHEA (50 mg/d), and 30 received placebo. Baseline lipoprotein profiles in the elderly were compared to healthy younger participants. Low-density lipoprotein (LDL) and high-density lipoprotein (HDL) particle sizes and concentrations were quantified using nuclear magnetic resonance spectroscopy. Results: The elderly had higher concentrations of total cholesterol, triglycerides, LDL cholesterol, total LDL particles, and small, dense LDL particles than the young. In men, neither DHEA nor testosterone affected LDL or HDL particle concentrations. In women, DHEA reduced HDL cholesterol [median difference (95% confidence intervals), Ϫ5.0 (Ϫ8.0, Ϫ2.0) mg/dl; P ϭ 0.002] and the number of large HDL particles [Ϫ1.0 (Ϫ1.8, Ϫ0.2) ␮mol/liter; P ϭ 0.003]. Conclusions: Long-term DHEA and testosterone had no significant effect on plasma lipoproteins in elderly men, but elderly women showed a lowering of the large HDL particles that may have potential adverse clinical implications. (J Clin Endocrinol Metab 95: 1617–1625, 2010)


orldwide, the segment of the adult population 60 yr of age and older is rapidly expanding due in part to a prolonged life expectancy. This rapid expansion of the aging population is associated with concomitant increases in the incidence of many age-related chronic diseases, including cardiovascular diseases (CVDs), which places an enormous burden on health care delivery and cost. Hence, it is vital to develop strategies to prevent or delay the onset

of these age-related chronic diseases. One such strategy may be to replace hormones that decline with age. There are many observational studies in humans (1, 2) and experimental studies in animals (3, 4) that link dehydroepiandrosterone (DHEA) and cardiovascular risk. DHEA and its sulfated ester (DHEAS) are the most abundant circulating steroid hormones in the human body (5). There is a progressive decline in DHEA concentrations in
Abbreviations: BMI, Body mass index; CAD, coronary artery disease; CI, confidence interval; CVD, cardiovascular disease; DHEA, dehydroepiandrosterone; DHEAS, sulfated form of DHEA; HDL, high-density lipoprotein cholesterol; IQR, interquartile range; LDL, low-density lipoprotein cholesterol; NMR, nuclear magnetic resonance.

ISSN Print 0021-972X ISSN Online 1945-7197 Printed in U.S.A. Copyright © 2010 by The Endocrine Society doi: 10.1210/jc.2009-2000 Received September 18, 2009. Accepted January 7, 2010. First Published Online February 5, 2010 * M.S. and B.A.I. contributed equally to this work.

J Clin Endocrinol Metab, April 2010, 95(4):1617–1625



CA). Being a sex steroid precursor hormone. CA). testosterone deficiency has been shown to increase the atherogenic potential of lipoprotein particles (7). 11). irrespective of their total concentrations of lipids and apolipoproteins (12). we hypothesized that DHEA replacement in healthy elderly men and women.g. fasting blood samples were collected at baseline and at 24 months for measurement of lipid and lipoprotein concentrations. April 2010. it is important to address the effectiveness and safety of DHEA or testosterone replacement in improving cardiovascular health on a long-term basis (18). 13). as previously reported (19).6 nmol/liter) for elderly men and a DHEAS concentration of less than 0.1618 Srinivasan et al. DHEA has been marketed as an “antiaging” supplement claimed to protect against CVD and other chronic illness.6 ␮mol/liter) for elderly women. a placebo tablet plus a transdermal testosterone patch (5 mg/d.. Although the clinical utility is not fully established. 2) and experiments in animals (3. In addition. Exclusion criteria included evidence of clinically important coexisting illness or conditions. Baseline characteristics are presented in Table 1 and have been reported previously (19). For example. participants received a new supply of tablets and patches from the Mayo Clinic’s Research pharmacy and were asked to return any unused tablets or patches to monitor study compliance. 16). For the present subanalysis. and Olmstead county area by local radio and newspaper advertisements. A total of 92 elderly men and 62 elderly women were randomly assigned to the study groups. DHEA Replacement and Lipoprotein Particle Sizes J Clin Endocrinol Metab. DHEA is thought to influence lipid metabolism and thereby CVD risk (6).. and 2) to evaluate the role of DHEA or low-dose testosterone replacement for 2 yr in elderly men with low DHEAS and bioavailable testosterone levels and DHEA in elderly women with low DHEAS on the lipoprotein profiles. Fasting blood samples were collected at baseline and every 3 months for the measurement of liver enzymes. two-site chemiluminescence immunometric assay (Immulite. hematocrit. Every 3 months. were chosen to ensure that a sufficient number of healthy elderly people could participate in the study. Experimental design Elderly men received either a DHEA tablet (75 mg/d) plus a transdermal placebo patch. placebo-controlled study done in the absence of industry support. would result in improvements in the lipid and lipoprotein profile. and DHEAS. Similar to DHEA. Bioavailable testosterone was measured on the basis of differential precipitation of SHBG by ammonium sulfate after the equilibration of serum samples with .95 ␮g/ml (2. families with exceptional longevity have been shown to have larger LDL and HDL particle sizes. lipoprotein particle sizes and concentrations assessed by nuclear magnetic resonance (NMR) have been shown to be superior to traditional lipid measures in predicting CVD risk (14). Lipoprotein particles have been shown to exhibit ageand sex-related characteristics (15). Whether replacement of DHEA or testosterone has any effect on reversing these age-related changes in lipoprotein particle profile is unknown.57 ␮g/ml (4. with large low-density lipoprotein (LDL) and large high-density lipoprotein (HDL) particles associated with reductions in CVD risk (10. It is known that females have less atherogenic lipoprotein profile compared with males (e. Mountain View. Los Angeles. as well as testosterone replacement in healthy elderly men. Alza Corp. there is also an inverse relationship between CVD risk factors and LDL and HDL particle sizes (12. Hence. specifically increasing their atherogenic potential. only the participants with a complete set (baseline and 24month) of unfrozen EDTA plasma samples were included. 9). Irrespective of sex. higher HDL concentrations) (15. Due to the widespread use of these hormones. SHBG was measured by solid-phase. lipoprotein particle sizes undergo undesired changes with age (17). Briefly. which represented the 15th percentile of levels for normal young men and women (20). 95(4):1617–1625 both men and women with advancing age (5). this is a 2-yr randomized. D-TRANS.3 ␮mol/liter) and a bioavailable testosterone concentration of less than 103 ng/dl (3. Therefore. In particular. which was necessary for the NMR lipoprotein profiling. testosterone. There is growing evidence indicating that lipoprotein particle size plays a role in mitigating CVD risk. A detailed description of the study design and methodology has already been described as part of another publication (19). or a placebo tablet plus a placebo transdermal patch. For young participants only baseline blood samples were collected. double-blind. Moreover. These cutoff values. Diagnostic Products Corp. Minnesota. Cross-sectional studies have also suggested that low levels of bioavailable testosterone may lead to increased risk of coronary artery disease (CAD) in men (8. Volunteers were eligible to participate in this study if they were 60 yr of age or greater and had both a DHEAS concentration less than 1. 4). MN). Elderly women received either DHEA (50 mg/d) or placebo. testosterone also declines with age in men (7). Based mainly on data from observational studies in humans (1. Subjects and Methods Subjects The study was approved by the Mayo Clinic Institutional Review Board (Rochester. the aim of the current study was: 1) to assess the impact of age and sex on lipoprotein particle concentration and size as assessed by NMR. Analytic procedures Plasma concentrations of DHEAS and total and bioavailable testosterone were measured by competitive chemiluminescence immunoassay. Potential research volunteers were recruited from the greater Rochester. Also included were 37 healthy young men and 38 healthy young women between the ages of 18 and 31 yr to obtain a baseline comparison of outcome measures. especially on LDL and HDL particle sizes and concentrations as measured by NMR.

Changes in the outcome measures were calculated by comparing the value at baseline with that taken at the last measurement. 5.9. 0.7) 0. 5. length of follow-up.4) 0. 16.21 (0. 0.3) 12.97) 390 (251. 31 of 32 subjects completed the placebo arm.5. 0.3.0 (3. Baseline demographics Table 1 summarizes the baseline characteristics for the participants included in the present analysis. 64. 10. 30 –100 pg/ml.5) 2. 43. 0.9) 0. Overall study compliance on average was above 95% for all treatments.5 (90. total testosterone—women 20 – 40 yr old.7) and 23.7 (56.53.9 (57. 0. 4.9 (15.9) DHEA 25 68 (66.3) 23.57 ␮g/ml. respectively.20) 398 (296.7) 2. the median duration of treatment with DHEA and placebo was 23. Nonparametric analyses were chosen in the present study because the lipoprotein particle data were not normally distributed.3.5. (Raleigh.8) 1. statin therapy was not included in the final model.1 (5.43) 30 (27.9. 27. 0. 5.7. respectively.7.3) 20. 0.6. testosterone. 16. For elderly women. and the baseline value.5) months.4 (39.7.9 (25.2) 7. and 23.3 (50. and 27 of 30 subjects completed the testosterone arm in the parent study (19).6. 10. 33) NA 7.1) 4.3 (58.1) 1. 28.5) 0.11.3) Variable label n Age (yr) BMI (kg/m2) Body fat (%) Visceral fat/total fat ratio FFM (kg) Fasting glucose (mg/dl) Fasting insulin (␮U/ml) Insulin sensitivitya C-peptide (nmol/liter) DHEAS levels (␮g/ml) Total testosterone (ng/dl) Bioavailable testosterone (ng/dl) Estradiol (pg/ml) Bioavailable estradiol (pg/ml) Placebo 30 70 (66.0.15) 38.1 (36. below 8%.0) 29.21 (0. and men 18 –30 yr old.6 to 23.4) 0. 64. and men.7.5) 89. and men. estradiol—premenopausal women. 0.1 pg/ml. 45.9. 29 of 30 subjects completed the DHEA arm. 19.org 1619 TABLE 1. Multiple regression analyses were conducted to examine the treatment effects on the change in each of the outcome variables.1)b 3. 465) 56.7. 128 – 430 ng/dl. 24. 5.14.2 to 23.3.2 (1. 72) 28.1 (1.1 (23.2 (0. the sex of the participant.2 (22.63 (0. Baseline comparisons between treatment groups were conducted using a nonparametric Wilcoxon rank sum test. below 6%.3) 92. 93.9.4) 42. 72) 27 (24.1) 3. 10.2–1.17. 1.1 (22. There were no differences noted between groups in terms of age.3.4 (25.J Clin Endocrinol Metab. DHEA). 16. the dependent variable was the change from baseline to 24 months (with the use of a rank transformation).05 (elderly females: placebo vs. respectively.6.2 (5.7) 1. 32.30. respectively.4) 12. 30. pravastatin.5) 0.21) 62.5) 0.6 (89.8) 9. tracer amounts of tritium-labeled testosterone.8 (1.5) 0. because there was no effect of statin therapy on the . NC).3 (6. and men 20 – 40 yr old. In the group of elderly men.2) Data are expressed as median (interquartile range).4) 9.30. the age of the participant at enrollment.8 (6. The first few analysis batches contained a mixture of elderly male. and 27 of 30 subjects completed the DHEA arm in the parent study (19).90) 357 (282.16) 39.0.5 pg/ml.7 to 23.2) DHEA 19 68 (65.5) 0.2.8).7) 8. 0. 1.1) Testosterone 23 66 (62. 30 of 30 subjects completed the placebo arm.2. 97. ratio of visceral fat to total body fat. total and bioavailable testosterone.6 (86.9. as described previously (21).0.77 (0.4 (6. Fat free mass.9 (7. 23.0 (25. 8. visceral fat. 5. below 8%. bioavailable estradiol—premenopausal women. a b Insulin sensitivity index calculated from an oral glucose minimal model as described previously (40).50. 42.7. 99. Normal ranges: DHEAS—women 18 –30 yr old. below 10%. Data maintenance.5) 27. NA. 11.8) 3. 30.1 (1.8. body mass index (BMI).4 to 23.9 (36. The lipoprotein particle size and concentrations were measured in plasma in the Cardiovascular Medicine Laboratory at the Mayo Clinic using a 400-MHz NMR lipoprotein analyzer system developed by LipoScience Inc.2 (89. The interassay coefficients of variation for DHEA.4.0) months. 30. 95(4):1617–1625 jcem. Samples were run one or two times per week in batches of 60 – 64 samples.18.3 (6. The intraassay coefficients of variation for DHEA. 65.7 to 24. or fat free mass. 98. in elderly women there were significant differences in BMI and Statistical analysis Baseline demographic characteristics are presented as the median with the interquartile range (IQR).6. 442) 62.1.0 (24. April 2010.7 (7.1. 24. elderly female.1.0. and the independent variables included were the treatment group. total body fat.6) 93. editing.40 (0. weight. bioavailable testosterone— men 20 – 40 yr old. 95. However.8 (46.6) 1. 0.08.44 –3. and placebo groups in elderly men was 23.6 –7.9 (3.7. 72) 27. Results The median duration of treatment within the DHEA.7) 93. 0.0 –35. 62.9 (89.7 (37. Wilcoxon rank sum P Ͻ 0.9. and SHBG were below 12%. and below 5%. 10 –55 ng/dl. 14.13 (0. 75) 27.0 pg/ml. not available. (0.1 (1.3.3 (52.endojournals. 465) 52. total and bioavailable testosterone.5 (6.6 (6. 69. 29.0 (22.9. 350-1230 ng/dl.7.0) 19.89 – 4. 1. lovastatin. Baseline characteristics of participants Elderly females Elderly males Placebo 29 66 (64. 12.9.7 (2.0) 8.5 (24. 0. 69) 26.2 (22. 40. and simvastatin).3 (22. In the group of elderly women.0) 0. 27. 28. 0.2 (0. Because approximately 25% of the elderly participants were taking statins (including atorvastatin.7.1)b 41.2.5) 94. young male.40.27) 59. 1. FFM.2. and analysis were carried out in the Division of Biostatistics at the Mayo Clinic.4) 11.73 (0. statin therapy was considered as a potential covariate.23) 59.0.5 (19.4.1 (21. 3.2 (2.32 ␮g/ml.0 (16. 1. 35) NA 8. change in any of the clinical outcomes in response to DHEA or testosterone.0.5).20 (0.7) 25. and below 6%.4.32 (0.1) 4.3 (3. and young female samples.53) 28 (27. 63. 1. and SHBG were below 7%.3) 1. However.

Treatment with DHEA increased the level of total testosterone by a median of 19. which would represent a partial restoration toward levels seen in young adults. 960) 21. 555) 337 (205.4) Variable label n LDL cholesterol (mg/dl) Baseline Final Total LDL particles (nmol/liter) Baseline Final Large LDL particles (nmol/liter) Baseline Final Medium-small LDL particles (nmol/liter) Baseline Final Very small LDL particles (nmol/liter) Baseline Final Total small LDL particles (nmol/liter) Baseline Final LDL mean particle size (nm) Baseline Final Placebo 30 108 (97. and total testosterone have been previously reported (19). 202) 527 (336. Fig. 221) 177 (104.3 (20. 95(4):1617–1625 fasting glucose at baseline between treatment groups. In brief.0. 224) 822 (300. As previously reported (22). April 2010. 127) 108 (98. 110) 98 (92. fasting insulin.5) DHEA 25 101 (88. 174) 411 (253.8 IU/liter and 3. 897) 679 (398. 196) 541 (317.0 IU/liter. 683) 532 (364.0) Data are expressed as median (interquartile range).4 (20. 108) 1042 (834. 21. 21. 1325) 1044 (869. fasting glucose concentrations.4. In contrast. 119) 97 (77.6. 22. 110) 1325 (928. 112) 93 (82.6 (20. 122) 1195 (984. treatment with testosterone raised total testosterone concentrations to levels seen in young adults and also raised bioavailable testosterone by approximately 50%. treatment with DHEA did not significantly affect total testosterone in elderly men relative to placebo. 437) 140 (82. 19 of the original publication of the results from this study demonstrates that the DHEA treatment resulted in elevations in DHEAS concentrations after 3 months of treatment and that these elevations were maintained for the duration of the TABLE 2. 613) 436 (368.1620 Srinivasan et al. In addition. 1303) 1126 (988. 176) 100 (64. 528) 201 (110. 780) 1029 (413. 21. 890) 586 (350. 582) 484 (349.5 ng/dl (3. 848) 21.6. 1498) 1205 (1039.6 nmol/liter) and bioavailable testosterone by a median of 30. 746) 700 (421. 749) 508 (321. 21.4 ␮g/ml (9. 678) 648 (408.6) 20.8) 21. 115) 1167 (876. 1125) 20. 1386) 452 (340. The incomplete restoration of bioavailable testosterone may be due in part to a testosterone-induced suppression of endogenous testosterone production as indicated by declines in the concentrations of both LH and FSH relative to placebo (Ϫ1.7 nmol/liter) in elderly women relative to placebo. 21. or insulin sensitivity index. 21. 197) 156 (94. bioavailable testosterone.1 (20. 1020) 609 (334.9) 21.4) 20. 1309) 1057 (828. No significant differences were observed between treatment groups at baseline.2.5 (20.3 (20.7 (20. 445) 288 (192. 1323) 1014 (821.7 (20.8. 821) 522 (295. Effect of DHEA or testosterone in elderly men and DHEA in elderly women on DHEA and testosterone concentrations The changes in DHEAS. 168) 128 (85.4) Testosterone 23 105 (91.8 (20.3 ␮mol/liter) in elderly women. respectively) (19). DHEAS values on therapy reached levels equivalent to that of high normal levels seen in younger adults.3. 112) 94 (82. study in both elderly men and elderly women.4) 20. 1097) 885 (516. 472) 147 (80. 21.7) DHEA 19 93 (85. 1403) 501 (336. 21. 480) 376 (269. 627) 104 (61. Treatment with testosterone increased the level of total testosterone by a median of 104. Baseline lipoprotein particle sizes and concentrations The baseline values for all LDL and HDL lipoprotein particle concentrations and sizes are summarized in Tables 2 and 3. Importantly. 1445) 304 (207. bioavailable testosterone (in men).4. respectively. there were no significant differences observed between groups for DHEAS.8 ␮g/ml (10. 767) 668 (422.4 ng/dl (1. 195) 121 (82. 2 in Ref. 965) 20. . 171) 511 (318.5. 123) 958 (827. 923) 744 (444.4 (20.8 ng/dl (0.2 ␮mol/liter) in elderly men and by 3. DHEA therapy increased the plasma levels of DHEAS by a median of 3. 1339) 401 (283. 664) 421 (232. 939) 21. total testosterone.5. 1328) 374 (264. Figure 2 from the original publication of the results from this study (19) also demonstrates that testosterone treatment resulted in elevations in bioavailable testosterone after approximately 6 –9 months of treatment and that these elevations were maintained for the duration of the study in elderly men. 821) 662 (426. DHEA Replacement and Lipoprotein Particle Sizes J Clin Endocrinol Metab. 21. 841) 490 (267. 627) 129 (87.1 nmol/liter) in elderly men relative to placebo. No major differences in lipoprotein Baseline and 24-month (final) LDL cholesterol profiles by sex and treatment Elderly females Elderly males Placebo 29 93 (83.

Both elderly men and women had higher total cho- .2) 9.3. 4. 9.5) 1.0) 2.6) 23.9. Both elderly men and women had higher total cholesterol (P Ͻ 0.7) 8.9) 6.5 (2. respectively.6 (8.7 (27.1) (25. 2.2) 29. This may be related to differences in sex hormone profiles (15.0) (3.5 (0. 26. 87 mg/dl (67.1) 23. at baseline the elderly women in the placebo treatment group had higher triglyceride concentrations than the DHEA group [median (IQR). size or concentration were noted (P Ͼ 0. total HDL particles. 35. total HDL particle. 31.4) 29.5.001).1. 107 mg/dl (87.5. 52) 47 (40. 9. P ϭ 0. LDL cholesterol.2) 0.6 (2.4. 34.3. triglyceride. and large HDL particle concentrations (P ϭ 0. HDL cholesterol. 6. 1).4.6 (8.0.8) despite significantly increasing DHEAS and testosterone concentrations. 6.0) 6. 3. and large HDL particle concentrations in elderly women. 6.8. 3) 2 yr of either DHEA or testosterone replacement had no measurable affect on the lipoprotein particle profiles in elderly men.2) 9. 9.7 (8.5. Discussion The main findings from the current study are: 1) elderly men and women had higher atherogenic potential of their lipoprotein particle profiles than their younger counterparts.6.113). 3. 25.6 (8.4.001).8) 0.5) 24.1) 8. 24.J Clin Endocrinol Metab.2) 1. 23.8. 2. 26. Total HDL cholesterol.7. 6. Effect of age and sex on lipoprotein particle sizes and concentrations Figure 1 shows the effect of age and sex on total cholesterol. 8. Two years of treatment with either DHEA or testosterone in elderly men did not significantly affect the LDL or HDL particle profiles (P Ͼ 0.0 (27. 4. 5. and small.8 (0. 1.9) 8. 43) DHEA 25 38 (34.5) 8. 54) 31. 8. 9.001). 1. 16).8. 30.4.9 (19. 41) 39 (35.7 (26.8) 1.0) 1.04) and elderly women (P ϭ 0.8) 3. In addition.4) 28. April 2010.6) 29.3 (17.3) 22.org 1621 TABLE 3. With regard to LDL. 27.0) 23.003) relative to placebo.7) 32. 9. total HDL particle.5. to levels found in young men.4 (19. In contrast.6 (8. 33.5) 22. 34.5 (0. 4. 31.1 (4. 55) 46 (34. 43) 39 (34.0) 0.4.05) between treatment groups at baseline within the elderly men. However. 27.1 (0. 8.3 (19.8 (2. 44) 36 (34.0 (8. 46) Testosterone 23 36 (30.4 (20. and large HDL particle concentrations were higher in women compared with men (P Ͻ 0.001).9 (0. 25.3.7.endojournals.5. 45) Variable label n HDL cholesterol (mg/dl) Baseline Final Total HDL particles (␮mol/liter) Baseline Final Large HDL particles (␮mol/liter) Baseline Final Medium HDL particles (␮mol/liter) Baseline Final Small HDL particles (␮mol/liter) Baseline Final HDL mean particle size (nm) Baseline Final Placebo 30 46 (40.1) 8.135) vs.0. 5. 9.4. the present data demonstrated that women had significantly higher concentrations compared with men of total HDL cholesterol.0) 2.4 (2.1.7) 5.03) than their younger counterparts.7 (2.2.001) and triglyceride (P Ͻ 0.5. and large HDL particles (Fig.0.7 (26. the elderly group had elevated LDL cholesterol (P ϭ 0. The current study supports previous data suggesting that an individual’s sex influences his/her lipoprotein profile.8) 9.05) concentrations compared with their younger counterparts.2 (0. The large HDL particles were lower in both elderly men (P ϭ 0.4) 28.0) 4.8 (0.4) 29.0 (1. dense LDL particle concentrations (P Ͻ 0. 9.5) 23. 31.4) 9.2.0 (8. 5.8 (28.1 (18.2 (21.3) 22.3 (19. 5.3 (19.002).4 (4.7 (27. total LDL particle (P Ͻ 0.1 (2. 55) DHEA 19 48 (42.2 (0.5) 21.7 (21.5) 3.3 (30.4.0) 8.7. 31.0 (8.5.0) Data are expressed as median (interquartile range). 9. and LDL and HDL particle concentrations.026].1) 19. Specifically. DHEA replacement did not alter the LDL particle profile in elderly women.4. 7. 9. 24.4 (2.3 (29.7 (0.05). (26. Effect of DHEA or testosterone in elderly men and DHEA in elderly women on lipoprotein particle sizes and concentrations Table 3 summarizes the treatment effect with DHEA compared with placebo in elderly women and DHEA or testosterone compared with placebo in elderly men on lipoprotein particle size and numbers. 95(4):1617–1625 jcem. No significant differences were observed between treatment groups at baseline. 2) 2 yr of DHEA replacement resulted in reductions of HDL cholesterol. there were also measurable effects of age on the lipoprotein particle profiles in both sexes as shown previously (15).9) 5. 32.4.7) 32. total HDL particle concentrations (P ϭ 0.4.6 (8. Two years of treatment with DHEA in elderly women significantly reduced total HDL cholesterol concentrations (P ϭ 0. Baseline and 24-month (final) LDL cholesterol profiles by sex and treatment Elderly females Elderly males Placebo 29 37 (34.4) 3.1 (8.

CVD risk factors. (interquartile range). dense LDL particles increased substan. In a study involving Ashtially in men. Effect of age and sex on various lipoprotein particles. it has been reported that lower levels of small LDL particles were artery disease (15). Females generally have a less atherogenic lipoprotein profile compared with males (15. Moreover. We observed that DHEA replacement in elderly men associated with better cardiovascular health and longevity had no measurable affect on the lipoprotein profiles. and large HDL particle concentrations in elderly replacement resulted in an unfavorable effect on lipoprowomen. is the leading cause of morbidity and mortality among adults in the United States. In randomized trials. and its incidence is increasing in women and with age (27). total HDL particle. The potential clinical implications that the DHEA-induced lowering of HDL cholesterol.1622 Srinivasan et al. (12). DHEA Replacement and Lipoprotein Particle Sizes J Clin Endocrinol Metab. substantial evidence lesterol and triglycerides compared with their younger is emerging about the relationship between lipoprotein counterparts. lipoprotein particles undergo several distinct modifications with age that make them more FIG. in particular CAD. independent of sex. and large HDL particle concentrations have on CVD outcomes among elderly women remain to be fully understood. Concentrations of large HDL and large LDL partiwhereas DHEA replacement significantly altered the HDL cles were significantly higher in individuals with exceplipoprotein profiles in elderly women. these lipid alterations may have been mediated by the DHEA-induced elevation in total testosterone that was observed among the elderly women receiving DHEA. but not elderly men. short-term (3– 6 months) replacethat indicate that exogenous DHEA administration results in reductions in HDL cholesterol in women (23–25). With advancing age. we also recently reported that short-term (3 months) DHEA supplementation (50 mg/d) in hypoadrenal women (characterized by absolute deficiency of DHEA) results in reductions in the large HDL particle concentration (26). Sim. It remains uncertain why the lowering of HDL cholesterol. It therefore is a concern that DHEA cle. 29). total HDL particle. 1. thus indicating a higher risk for coronary kenazi Jewish centenarians and their offspring. However. Lipoprotein particle sizes as estimated by NMR spectroscopy have recently emerged as an important tool in the risk stratification for CAD (14).ment of DHEA has been shown to enhance cardiometa- . April 2010. and large HDL particle concentrations occurs in response to DHEA replacement in elderly women. the concentrations of atherogenic small. thereby increasing CAD risk in both sexes. These results are consistent with previous reports tein particles in women. However. CVD.particle sizes and longevity. 16). Data are presented as median susceptible to oxidation (28. Specifically. 95(4):1617–1625 ilarly. All P values are based on nonparametric Wilcoxon rank sum tests. DHEA tional longevity (13) and had an inverse relationship with significantly reduced HDL cholesterol. total HDL parti.

1) Ϫ0.0 (Ϫ3.0) Ϫ0.9) 0. dyslipidemia. 0.774 0. The route of administration (transdermal patch vs.099 0. 0. Therefore.611 0.5) 0.0 (Ϫ17. However. Specifically.0.20) Ϫ4.1) 0.844 0. There is conflicting evidence regarding testosterone replacement on cardiometabolic parameters in randomized studies as indicated by a recent meta-analysis (34). 68.0.2. and the metabolic syndrome (32.20) 0. As stated previously. and glucose tolerance in elderly adults with low androgen levels (19).0 (Ϫ77.0.0 (Ϫ114.422 0.443 42.7 (Ϫ2.5) 0. Ϫ0. 1.0 (Ϫ89.118 0.0) 2.737 1.9.004 0.0) 9.0) Ϫ0. 38.5 (Ϫ65. 5. 0.7.30. these beneficial effects on insulin sensitivity and visceral adiposity have not been shown to be sustained over longer periods (2 yr) of treatment (19). 35).0 (Ϫ1. duration. It is noteworthy that in the present study low-dose testosterone (5 mg/d) was used to minimize the risk of adverse effects like prostate enlargement and cancer.9) Ϫ0. 0.988 0.1. and baseline values.6) Ϫ1. sex. 34.0 (Ϫ0.0 (Ϫ91.2 (Ϫ0. it is not likely that our P value Data are expressed as median difference (95% CI).5 (Ϫ4.0) 13.0) Ϫ2.05 (Ϫ0. All P values are two-sided and are based on multiple regression analysis in which the dependent variable was the change from baseline (with the use of a rank transformation) and the independent variables were the study group.0) Ϫ29. placebo The impact of DHEA and testosterone on LDL and HDL cholesterol profiles P value Elderly females DHEA vs. 106. 4.0) 48.984 0.10 (Ϫ0.0.0) 38.0) Ϫ19.927 0. 3.3.632 0.0.5 (Ϫ96.8.7 (Ϫ0. April 2010.10 (Ϫ0.0.0.1 (Ϫ0.0 (Ϫ1.0 (Ϫ10.732 0.0. 60.40. and importantly the route of administration (oral vs.0) 8. 167. 205. placebo TABLE 4.0 (Ϫ20.0) 43.0) Ϫ5.243 0. an a priori power analysis was not feasible for the present outcomes.0) 48. 186.279 0.9) 1. 143. safety and efficacy for longer duration of its use remains uncertain.0 (Ϫ9. Although there are number of studies that have examined the impact of testosterone replacement on lipid profile (7. 0.0.799 0. length of follow-up.2) Ϫ0. Therefore.2.0. 0.5 (Ϫ1.6.40. 0. Testosterone vs. 180.0.0. 3.org 1623 bolic parameters by improving insulin sensitivity and reducing visceral adiposity (30.358 Elderly males P value DHEA vs.0 (Ϫ98.502 0. This is probably due to a difference in doses. 0. 12. In the current study.1. However. 2.8) 0. visceral adiposity. 0. it could be argued that these beneficial effects of DHEA replacement on insulin sensitivity and visceral adiposity outweigh the negative effects of DHEA on the lipoprotein particle profile.603 0. 31).0) 0. 142. 8.0.670 0.0) 0.0 (Ϫ30.0.0) 0.0 (Ϫ121. as previously reported (19).0 (Ϫ115. physical performance. the present study is a subanalysis of a larger clinical trial that was initially powered to detect significant effects of long-term supplementation with DHEA in elderly men and women and testosterone in elderly men on body composition. bone mineral density.347 0.966 0.261 0.0. Variable label LDL cholesterol LDL cholesterol (mg/dl) Total LDL particles (nmol/liter) Large LDL particles (nmol/liter) Medium-small LDL particles (nmol/liter) Very small LDL particles (nmol/liter) Total small LDL particles (nmol/liter) LDL mean particle size (nm) HDL cholesterol HDL cholesterol (mg/dl) Total HDL particles (␮mol/liter) Large HDL particles (␮mol/liter) Medium HDL particles (␮mol/liter) Small HDL particles (␮mol/liter) HDL mean particle size (nm) Ϫ5.0.0) 19.5732 0.290 0.844 0.0 (Ϫ8. 1. 2.002 0.0 (Ϫ68.5) 0. 33).758 0.endojournals. 0. 2 yr of testosterone replacement had no measurable affect on the lipid and lipoprotein profile.3. It is known that serum concentrations of testosterone decline with age in elderly men (7).691 0.339 0. 29.0.0 (Ϫ100.2) 0.0 (Ϫ12. Ϫ0.7) Ϫ0. 2.1 (Ϫ0.0. Ϫ2.0) Ϫ0.0 (Ϫ114. thereby increasing CVD risk. on the basis of the 95% confidence intervals (CIs) shown in Table 4.0) 0.687 0.0 (Ϫ1. and testosterone deficiency has been shown to increase the predisposition to insulin resistance.0.862 0. The present findings are in contrast to previous reports that indicate that testosterone administration at supraphysiological doses decreases HDL cholesterol in healthy young and elderly adults (36 –38). im vs.10) 0.0 (Ϫ0. we have previously reported from this same cohort that 2 yr of DHEA replacement had no measurable effects on either insulin sensitivity or visceral adiposity (19).971 2.961 0.J Clin Endocrinol Metab.0.003 0. data regarding the impact of testosterone on lipoprotein particle sizes and concentrations is still limited.3 (Ϫ1.2 (Ϫ0. age at the time of randomization.591 0. 95(4):1617–1625 jcem.3 (Ϫ3. 126. 160.248 0.237 .0) 27.0) Ϫ49.8. im injection) may also play a role in mitigating the effects of testosterone on cardiometabolic parameters.0) 36. placebo Ϫ1. transdermal) employed in different studies. Although a higher replacement dose (200 mg/every 2 wk) for shorter duration (3 months) has been shown to improve cardiometabolic parameters (39).0.1.2 (Ϫ2.

95(4):1617–1625 negative findings can be attributed to the small number of subjects in the study. Otvos JD.gov number NCT00254371. April 2010. Sulcova J. Similarly. Vogelman JH 1984 Age changes and sex differences in serum dehydroepiandrosterone sulfate concentrations throughout adulthood.A. Starka L 2000 Do DHEA/DHEAS play a protective role in coronary heart disease? Physiol Res 49(Suppl 1):S43–S56 7. Ajisaka R 2008 Testosterone and DHEA activate the glucose metabolism-related signaling pathway in skeletal muscle. and the NIH Roadmap for Medical Research.gov/aboutroadmap. Rubens MB. Barrett-Connor E. Rochester. Johannes CB. Riad-Fahmy D 1993 Endogenous sex hormones and ischemic heart disease in men. mortality. Acknowledgments The authors thank Bobbie Soderberg for her assistance with sample collection and storage. is supported by the W. both in elderly men and women.) and UL1-RR-024150-01 from the National Center for Research Resources (NCRR). Khaw KT. sree@mayo. Mohr BA. Christiansen C 2004 The aging male: testosterone deficiency and testosterone replacement. Steeds RP. SW. Rizer RL. Minnesota. Atherosclerosis 173:157–169 8.I. N Engl J Med 315:1519 –1524 3. Joseph 5-194. M. E-mail: nair.S. K. English KM. Gordon GB. Long-term (2 yr) DHEA replacement in elderly women resulted in a lowering of the HDL cholesterol. Heijmans BT. This work was presented at the American Heart Association’s 49th Cardiovascular Disease Epidemiology and Prevention Conference. Diver MJ. For example. Finally. adding statin therapy as a covariate had no affect on change in the individual lipoprotein particles. the 95% CI in the testosterone group for the changes in total LDL particle concentrations in elderly men ranged from Ϫ114 to 143 nmol/liter. Minnesota 55905. It should be recognized that approximately 25% of the elderly participants in the present study were taking statins at baseline. Mandour O. Clinical Trials. As we have previously reported (19). Circulation 106:3143–3421 12. the 95% CI in the DHEA group for the changes in total LDL particle concentrations in elderly men ranged from Ϫ114 to 106 ␮mol/liter (ϳ10% reduction to ϳ10% increase). In summary.nih.nih. Cobain MR. McKinlay JB 2001 Low dehydroepiandrosterone and ischemic heart disease in middle-aged men: prospective results from the Massachusetts Male Aging Study. M. Sweetnam PM. Similarly. J Clin Invest 82:712–720 5. Importantly. Araujo AB. Sato K. Address all correspondence and requests for reprints to: Dr. L. In contrast. the 95% CI in the DHEA group for the changes in total HDL particle concentrations in elderly men ranged from Ϫ3 to 3 ␮mol/liter (ϳ8% reduction to ϳ8% increase). Moreover.edu. Underwood SR. a component of the NIH. Diabetes 51:1949 –1956 11. A study in the hypercholesterolemic New Zealand white rabbit with aortic intimal injury. Disclosure Summary: The authors have no conflicts of interest and financial disclosure.D. Channer KS 2000 Men with coronary artery disease have lower levels of androgens than men with normal coronary angiograms. Yen SS 1986 A prospective study of dehydroepiandrosterone sulfate.ncrr. the 95% CI in the testosterone group for the changes in total HDL particle concentrations in elderly men ranged from Ϫ1 to 3 ␮mol/liter (ϳ3% reduction to ϳ10% increase). Aizawa K. and cardiovascular disease. Eur Heart J 21:890 – 894 9. Am J Epidemiol 153:79 – 89 2. DHEA Replacement and Lipoprotein Particle Sizes J Clin Endocrinol Metab. Mayo Clinic. 2002 Third Report of the National Cholesterol Education Program (NCEP) Expert Panel on Detection. Jones TH. and large HDL particle concentrations. Alexandersen P. Beswick AD. Evaluation. DHEA replacement increased the plasma levels of DHEAS to levels typically observed in younger adults. Fuller JH 2002 Lipoprotein subclasses and particle sizes and their relationship with coronary artery calcification in men and women with and without type 1 diabetes. Am J Physiol Endocrinol Metab 294:E961–E968 4. Brind JL. the participants in the present study were randomly assigned to receive either of the treatments or a placebo. DHEA replacement had no measurable effects on the lipoprotein particle profile in elderly men. and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III) final report. Houwing-Duistermaat JJ. Yarnell JW. Florida. Information on Reengineering the Clinical Research Enterprise can be obtained from http://commonfund. Colhoun HM. References 1. Ph. Arterioscler Thromb 13:517–520 10. Among the elderly population in the greater Rochester. There are also no industry relationships to disclose. Orentreich N.asp. and in elderly women ranged from Ϫ98 to 186 ␮mol/liter (ϳ10% reduction to ϳ20% increase). the CI would have been even narrower.1624 Srinivasan et al. these 95% CIs indicate that the number of participants was sufficient to establish any clinically meaningful treatment effects. Feldman HA. J Clin Endocrinol Metab 59:551–555 6. total HDL particle. Iemitsu M. Weisman HF 1988 Reduction of atherosclerosis by administration of dehydroepiandrosterone. 200 First Street .gov/. Sreekumaran Nair. Taskinen MR. This work was supported by National Institutes of Health (NIH) Grants AG-PO114383 and RO1 AG-09531 from the National Institute of Aging and Grants KL2 RR084151 (to B. Stephenson Fellowship. An up-date. it is difficult to find people who are not treated with diet and/or therapeutic agents when their LDL cholesterol concentrations are above 100 mg/dl. and Olmstead county areas. Porsova-Dutoit I.D. in Palm Harbor. The Caerphilly Prospective Study. Information on NCRR is available at http://www. If the study had enrolled more subjects. Longcope C. March 11–14. Beekman M. Bush DE. thus ruling out reductions in the LDL particle concentrations greater than 114 nmol/liter (ϳ8% reduction) or increases in LDL particle concentrations greater than 143 nmol/liter (ϳ10% increase).. long-term physiological replacement of testosterone had no measurable effects on the lipoprotein particle profile in elderly men. Its contents are solely the responsibility of the authors and do not necessarily represent the official view of NCRR or NIH. However. 2009.

Freedman DS. Goodwin E. Blazer D 2004 Committee on Assessing the Need for Clinical Trials of Testosterone Replacement Therapy. Hamdy O. Yen SS 1994 Effects of replacement dose of dehydroepiandrosterone in men and women of advancing age. Circulation 119: 480 – 486 Khalil A. Eur J Endocrinol 154:899 –906 Dalla Man C. Nadeau A. Short KR. McDermott M. Santanna J. Dangoisse V. 28.J Clin Endocrinol Metab. Cupples LA. and the development of type 2 diabetes in middle-aged men: prospective results from the Massachusetts Male Aging Study. Nelson JC. O’Donnell C. D’Agostino RB. Rizza RA. Clin Lab Med 26:847– 870 Henderson GC. Diabetes Care 25:55– 60 Stellato RK. Dhatariya K. Haase N. Klaus KA. McKinlay JB 2000 Testosterone. Ford E. Atherosclerosis 160:399 – 406 Khalil A. Steinberger J. Sacco R. Cupples AL. Nehra A. Kissela B. 20. Klee GG. Strom BL 2001 Effect of transdermal testosterone treatment on serum lipid and apolipoprotein levels in men more than 65 years of age. 37. Rivier JE. 1–10 Nair KS. Blauw GJ. Holmes JH. Bigelow ML. Fortin JP. Schaefer EJ. Stafford R. Gianfrilli D. 40. Lamarche B. Nair KS 2005 Effect of dehydroepiandrosterone replacement on insulin sensitivity and lipids in hypoadrenal women. Greco EA. 32. Cullberg G. Peachey H. Am J Med 111:255–260 Meriggiola MC. Rizza RA. Toffolo G. Cobelli C 2002 The oral glucose minimal model: estimation of insulin sensitivity from a meal test. Jensen MD 2006 DHEA in elderly women and DHEA or testosterone in elderly men. Hong Y 2009 Heart disease and stroke statistics—2009 update: a report from the American Heart Association Statistics Committee and Stroke Statistics Subcommittee. Melton 3rd JL. Dalla Man C. van der Ouderaa F. FEBS Lett 435:153–158 Khalil A. PLoS Med 3:e495 Barzilai N. Wedick NM. Powell J. Clin Chem 50:1189 –1200 Pascot A. Wasserthiel-Smoller S. Nolan JJ. Prud’homme D. Wilson PW. Sorlie P. Isidori A. 23. 38. Bremner WJ 1995 Testosterone enanthate at a dose of 200 mg/week decreases HDL-cholesterol levels in healthy men. Robins SJ 2006 Low-density lipoprotein and high-density lipoprotein particle subclasses predict coronary events and are favorably changed by gemfibrozil therapy in the Veterans Affairs High-Density Lipoprotein Intervention Trial. IEEE Trans Biomed Eng 49:419 – 429 . McConnell JP. Dhatariya K. Jensen MD. Diabetes 54:765–769 Villareal DT. Berman N. J Lipid Res 41:1552–1561 Liverman C. Basu A. Nair KS 2009 Higher muscle protein synthesis in women than men across the lifespan. Meigs J. Basu R. Vittone JL. JAMA 290: 2030 –2040 Otvos JD. Wingard DL 2002 Endogenous sex hormones and the development of type 2 diabetes in older men and women: the Rancho Bernardo study. 39. Hannoush P. Diabetes Care 23:490 – 494 Isidori AM. De Simone G. Shirazi A. Klee GG. Rader D. J Clin Endocrinol Metab 94:761–764 Lloyd-Jones D. Bergeron J. 22. Flegal K. Thom T. bone metabolism and serum lipid profile in middle-aged men: a meta-analysis. Westendorp RG. Callegari C. Wong N. Basu R. placement on lipoprotein profile in hypoadrenal women. O’Brien P. Hartman SJ. Clin Endocrinol (Oxf) 63:280 –293 Snyder PJ. Smith GE. 35. 15. Phillips J. 25. Samsioe G 1980 Administration of dehydroepiandrosterone enanthate to oophorectomized women– effects on sex hormones and lipid metabolism. Feldman HA. Riggs BL 1998 Relationship of serum sex steroid levels and bone turnover markers with bone mineral density in men and women: a key role for bioavailable estrogen. Fabbri A 2005 Effects of testosterone on body composition. Ho M. Klaus KA. 18. Greenlund K. JAMA 292:2243–2248 Oh JY. Fu ¨ lo ¨ p Jr T 1996 Increased susceptibility of low-density lipoprotein (LDL) to oxidation by ␥-radiolysis with age. Roger V. Atzmon G. 36. Ann Intern Med 116:967–973 Kapoor D. Barrett-Connor E. Shalaurova I. Lemieux I. 16. O’Brien P. Lenzi A. Tindall DJ. Nichol G. Marelli A. Channer KS. Ferguson TB. Bloomfield HE. O’Fallon WM. J Clin Endocrinol Metab 78:1360 –1367 Mortola JF. 31. sex hormone-binding globulin. Ford GC. Otvos JD.endojournals. Clevenger B. Paulsen CA. Schechter C. Fu ¨ lo ¨ p Jr T 1998 Age-related increased susceptibility of high-density lipoproteins (HDL) to in vitro oxidation induced by ␥-radiolysis of water. Collins D. Yen SS 1990 The effects of oral dehydroepiandrosterone on endocrine-metabolic parameters in postmenopausal women. Jay-Gerin JP. Khosla S. Loh L. Irving BA. Cromwell WC. McNamara JR. Schaefer EJ. April 2010. LeHoux JG. N Engl J Med 335:1–7 Bagatell CJ. 21. Maturitas 2:301–309 Srinivasan M. 14. Giannetta E. Melton 3rd LJ. Jones TH 2006 Testosterone replacement therapy improves insulin resistance. Shuldiner AR 2003 Unique lipoprotein phenotype and genotype associated with exceptional longevity. Rosamond W. Bonifacio V. Parise H. Otvos JD 2006 Lipoprotein particle analysis by nuclear magnetic resonance spectroscopy. FEBS Lett 392:45– 48 Dhatariya K. Vale WW. Dhatariya K. Mozaffarian D. Executive summary. FASEB J 23:631– 641 Morales AJ. Furie K. Usher D. Dlewati A. J Clin Endocrinol Metab 71:696 –704 Mattson LA. visceral adiposity and hypercholesterolaemia in hypogonadal men with type 2 diabetes. Storer TW. Lackland D. 33. Horton ES. Tremblay A. Howard V. Nair KS 2009 Effect of dehydroepiandrosterone re- 27. 95(4):1617–1625 jcem. Shalaurova I. Marcovina S. Tangkeo P. Schaefer EJ 2004 Sex and age differences in lipoprotein subclasses measured by nuclear magnetic resonance spectroscopy: the Framingham Study. Adams R. Couillard C. 17. Khosla S. Lisabeth L. Holloszy JO 2004 Effect of DHEA on abdominal fat and insulin action in elderly women and men: a randomized controlled trial. Lipton R. Lacombe G. N Engl J Med 355:1647–1659 Khosla S. 19. Wagner JR. Jeyarajah EJ. Hailpern S. 30. and failure of androgen administration to amend age-related decrements. 26. Atkinson EJ. Bremner WJ 1992 Physiologic testosterone levels in normal men suppress high-density lipoprotein cholesterol levels. Int J Androl 18:237–242 Bhasin S. Fu ¨ lo ¨ p T 2000 Age-related decrease of dehydroepiandrosterone concentrations in low density lipoproteins and its role in the susceptibility of low density lipoproteins to lipid peroxidation. Circulation 113:1556 –1563 Freedman DS. Tricker R. Despre ´ s JP 2002 HDL particle size: a marker of the gender difference in the metabolic risk profile. Bunnell TJ. J Clin Endocrinol Metab 83: 2266 –2274 Jeyarajah EJ. Caumo A. WylieRosett J. Washington DC: National Academy Press. 29. Knopp RH. Cobelli C. glycaemic control. Slagboom PE 2006 Lipoprotein particle profiles mark familial and sporadic human longevity. Go A. Casaburi R 1996 The effects of supraphysiologic doses of testosterone on muscle size and strength in normal men. Carnethon M. Cheng S. Zador G. 34. 24. Berlin JA.org 1625 13. Kittner S.