This Article Abstract Full Text (PDF) Submit a related Letter to the Editor Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Copyright Permission Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Liu, P. Y. Articles by Handelsman, D. J. Search for Related Content PubMed PubMed Citation Articles by Liu, P. Y. Articles by Handelsman, D. J. Related Collections Male Endocrinology

Age-Related Changes in Serum Testosterone and Sex Hormone Binding Globulin in Australian Men: Longitudinal Analyses of Two Geographically Separate Regional Cohorts

Department of Andrology (P.Y.L., D.J.H.), Concord Hospital and ANZAC Research Institute, University of Sydney, Sydney, New South Wales 2139, Australia; Department of Endocrinology and Diabetes (J.B., P.J.L.), Royal Perth Hospital, Perth 6847, Western Australia, Australia; Bone Research Program (C.M., M.J.S.), ANZAC Research Institute, University of Sydney, Sydney 2006, Australia; Division of Endocrinology, Diabetes, and Clinical Nutrition (C.M.), University Hospital Basel, 4031 Basel, Switzerland; Bone and Mineral Research Program (T.V.N., J.R.C., J.A.E.), Garvan Institute of Medical Research, St. Vincent’s Hospital and University of New South Wales, Sydney 2010, Australia; and Laboratory for Cancer Medicine (P.J.L.), Centre for Medical Research, School of Medicine and Pharmacology, University of Western Australia and the Western Australian Institute for Medical Research, Perth 6000, Australia

Address all correspondence and requests for reprints to: David J. Handelsman, MBBS, Ph.D., FRACP, Department of Andrology, Concord Hospital, ANZAC Research Institute, University of Sydney, Sydney, New South Wales 2139, Australia. E-mail: djh{at}anzac.edu.au.

	Abstract

Top Abstract Introduction Subjects and Methods Results Discussion References

 
Background: Cross-sectional studies from different populations show a variable decline in blood testosterone concentrations as men age. Few population representative cohorts have been followed up over time.

Objective: The objective of the study was to quantify longitudinally the change in serum testosterone and SHBG concentrations with age in two well-defined, representative but geographically widely separated regional Australian cohorts.

Subjects and Setting: The Busselton cohort comprises individuals aged 18–90 yr residing in Western Australia assessed prospectively since 1981. Sera were assayed from 910 men, from whom further samples were available 14 yr later in 480. The Dubbo cohort involves individuals aged 61–90 yr living in Eastern Australia. Baseline sera were collected from 610 men and additional sera on a second (n = 370) and third (n = 200) occasion from 1989 to 2004. Men from both cohorts are community dwelling and of predominately European origin.

Results: Longitudinal analyses show the following: 1) total testosterone declines comparably (P > 0.9) by 1.3% (Busselton) and 0.9% (Dubbo) per annum with the same rates of decline when analyses were restricted to men older than 60 yr of age; 2) annual changes in SHBG were also very similar in age-restricted analyses (2.3% vs. 2.5%, P = 0.48); and 3) the annual increase in SHBG was steeper in middle-aged and older men (P < 10^–3 vs. young men). These longitudinal changes were all up to 4-fold greater in magnitude, compared with cross-sectional analyses of baseline data.

Conclusion: In two separate regional Australian populations, blood testosterone fell and SHBG increased comparably with age. Age-related changes in blood testosterone and SHBG previously described in urban-dwelling men are the same in men who reside in smaller regional cities of another continent.

	Introduction

Top Abstract Introduction Subjects and Methods Results Discussion References

 
INCREASING POPULATION longevity creates a need to foster methods to promote healthy, independent, and enjoyable living. Reduced testosterone production could contribute to aging-associated increases in frailty, falls, and fractures and decreases in bone strength, muscle mass and strength, and physical performance (1). Whether the age-related decline in systemic testosterone exposure, described thus far in longitudinal studies from the United States (2, 3, 4, 5, 6) and Europe (7), occurs at a similar rate in other populations is not yet known. This would support the extrapolation of findings from any large-scale, randomized, controlled trials to test a testosterone intervention to different populations (1).

Representative population-based cross-sectional studies of ethnically comparable but geographically distinct cohorts have generally shown that total testosterone (TT) declines in adults with each year of age (8, 9, 10, 11), with few exceptions (12). Furthermore, the quasilongitudinal effect size estimates obtained have varied by as much as 10-fold from 0.1% per annum (10) to 1.6% per annum (13). This large variation in effect size estimates suggest that the age-associated changes in systemic testosterone exposure may differ among ethnically comparable but geographically distinct cohorts. Longitudinal studies are needed to resolve this. However, longitudinal studies are limited in number (2, 3, 4, 5, 6, 7); all examined men residing in urban populations, two involved less than 100 men (5, 6), and another two examined extremes of adult age (2, 7). Potential geographical differences in the longitudinal rate of these declines and whether such changes even occur in regional or rural locations have not been investigated previously. Theoretically, this could occur due to putative differences in nutrition or obesity in urban compared with regional centers. This is pertinent because obesity is the most important identified confounder of the age-related decline in testosterone because it amplifies (2, 3) but does not fully explain this phenomenon. Furthermore, geographical differences in the age-related decline of testosterone have been reported (14), although small sample size and assessment of salivary rather than blood testosterone are important limitations. Additionally, whether the fall in testosterone occurs consistently among young, middle-aged, or older men is underexplored, and this information may be important for the timing of potential interventions.

The present study therefore aimed to determine whether: 1) blood testosterone concentrations decline with age in two separate regional city cohorts located at opposite sides of the Australian continent; 2) the magnitude of these changes might differ by geographical region; 3) the rate of decline in testosterone varies among young, middle-aged, or older men; and 4) age-related changes are due to obesity.

	Subjects and Methods

Top Abstract Introduction Subjects and Methods Results Discussion References

 
Subjects

Participants were recruited from two population-based regional Australian centers from the west and east coast of the continent. The Busselton Population Survey is an ongoing representative and comprehensive survey of the cardiovascular and respiratory health of residents in the Shire of Busselton in regional southwestern Western Australia (15). Of the total adult male population of 2770 in 1981, 1458 subjects completed the survey. The present analysis was restricted to men for whom blood testosterone concentrations were measured from blood stored in 1981 (n = 910), which represents approximately 35% of the total adult male population in the Busselton shire. Of these men, 53% (n = 480) were reassessed in 1994–1995.

The participants from Dubbo are part of the ongoing Dubbo Osteoporosis Epidemiology Study (16). Briefly, since 1989, all men and women 60 yr old or older living in Dubbo, a regional inland city of 32,000 people in the state of New South Wales, Australia, were invited to participate. By July 2004, 862 men were participating in the study. The present analysis was restricted to men who had a baseline serum sample available (n = 610, 70.6%, defined as the first visit at which serum was collected). Thereafter men were followed up at approximately 2-yr intervals until 2004. About 60% (n = 370), 30% (n = 200), and 5% (n = 30) had a second, third or fourth blood sample available for analysis.

Each study was approved by its respective ethics review committees, and all subjects provided written informed consent. Consent included approval to store serum for future hormone or other measurements.

Laboratory measurements

For the Busselton cohort, all baseline blood samples were obtained between 0710 and 1240 h. All subsequent blood samples were collected early in the morning. Serum samples were stored at –20 C. TT and SHBG were measured by commercial chemiluminescent immunoassay (Immulite 2000; Diagnostic Products Corp., Los Angeles, CA.) in 2001 (for initial blood samples collected in 1981) and 2003 (for subsequent blood samples collected in 1994–1995). All samples were measured on the same platform by the same technician. Prior reports suggest that blood SHBG or testosterone concentrations do not change appreciably after samples are frozen at –20 C for decades (4). Coefficients of variation averaged 15.5 (TT) and 2.9% (SHBG).

For the Dubbo cohort, blood samples were collected since 1989 at a convenient rather than a fixed time of day, but most were collected in the morning. Serum samples were stored at –80 C until analysis in late 2004. TT and SHBG were measured by commercial immunoassays (DELFIA; Wallac Oy, Turku, Finland). Coefficients of variation averaged 10.3 (TT) and 7.9% (SHBG). All samples were assayed at one time.

For both cohorts, free testosterone and bioavailable testosterone were calculated (17, 18).

Statistical analysis

To minimize the effect of outliers, nonparametric methods were used. Kruskal-Wallis and signed-rank tests and Spearman’s rank-order correlation were performed as appropriate. P < 0.05 was construed as significant, although recognizing the multiple testing may violate nominal significance levels. Data are median (interquartile range).

Longitudinal analyses

For each individual with serial data, a slope term was calculated by linear regression as an annual change in that analyte. Both the absolute and relative (percentage) annual changes were calculated and separately analyzed. Within each cohort, significant annualized change in each analyte was evaluated by signed-rank test of nonzero slope. A difference in the relative change in testosterone or SHBG between the Busselton and Dubbo cohorts was assessed by Mann-Whitney U test. The main approach was to evaluate relative changes in these dimensionless slope parameters to minimize the impact of using different assay methods.

The effect of age on these relationships was examined by dividing men from the Busselton cohort into three groups aged 18–40, 40–60, and older than 60 yr. Examining differences among the three epochs of age was assessed by Kruskal-Wallis test without excluding outliers. Analysis of covariance (ANCOVA) was used to simultaneously assess the effects of continuous [age, body mass index (BMI)] and categorical (geographical location) variables. Because ANCOVA can be heavily influenced by outliers, 25 men with baseline TT less than 5 nM were excluded beforehand (<5% of all data).

Cross-sectional analyses

Cross-sectional analyses were also performed to allow comparison with previous publications. Linear regression was performed to ascertain the magnitude of the relationship between biochemical or anthropometric parameters with age. The slope is the annual change in that parameter or the change per year of age. This was converted to a relative (percentage) change per year of age by dividing the slope by the corresponding cohort median.

	Results

Top Abstract Introduction Subjects and Methods Results Discussion References

 
Subjects

Compared with the Dubbo (n = 610) cohort, men in the Busselton (n = 910) cohort were younger [51 (29) vs. 72 (8) years, median (interquartile range), P < 0.001], with a wider age range (18–90 vs. 61–90 yr), slightly taller [174 (9) vs. 172 (9) cm, P < 0.001] and had been followed for longer [12.4 (13) vs. 4.0 (6.1) yr, P < 0.001]. Weight [77 (14) vs. 77 (18) kg] and BMI [25.7 (4.4) vs. 26.0 (5.0) kg/m²] did not significantly differ.

Longitudinal analyses

Significant relative (and absolute) decreases in TT were evident in the Busselton cohort and increases in SHBG and BMI were evident in both cohorts: Table 1. The relative decrease in TT in the Dubbo cohort was not significant (P = 0.07).

In the three age groups from the Busselton cohort (Table 2), TT and calculated free testosterone did not differ by age group. However, the relative increase in SHBG in the Busselton cohort rose more steeply with age (P < 10^–4, Kruskal-Wallis test). Calculated bioavailable but not free testosterone also differed by age (P = 0.038 and 0.13, respectively, Kruskal-Wallis test). Bonferroni-adjusted post hoc analyses showed that this was due to significant (P < 0.05) differences between the youngest (aged 18–40 yr) and oldest (aged older than 60 yr) groups for both SHBG and calculated bioavailable testosterone. These findings were confirmed by two-way ANCOVA using a model to examine simultaneously the effect of age, BMI, and geographical location (data not shown). This analysis also showed that obesity (BMI) did not explain any of these relationships (P > 0.4 for BMI term).

TT fell (–0.32 ± 0.07%, mean ± SEM, P < 0.001) and SHBG increased (0.46 ± 0.11%, P < 0.001) per year of age in men from Busselton. TT exhibited a similar but not significant decline (–0.37 ± 0.37%, P = 0.08) per year of age in men from Dubbo whereas SHBG increased (1.10 ± 0.34%, P < 0.001). Analogous relationships were obtained with cross-sectional analysis of data from the next visit after baseline samples (not shown). The magnitude of longitudinal hormonal estimates exceeded those obtained from the cross-sectional analyses of the baseline data by as much as 4-fold (compared with relative changes obtained from longitudinal analysis shown in Table 1).

	Discussion

Top Abstract Introduction Subjects and Methods Results Discussion References

 
We show in two longitudinal population-based representative cohorts of men from different regional city cohorts at opposite geographical areas of Australia that blood testosterone falls and SHBG increases with age. Separately, we show in one of the few direct comparisons between separate cohorts that the magnitude of these changes are similar, even after adjustment for obesity (1, 2, 3). Obesity was examined as a confounder because of its immediate relevance due to the escalating global obesity epidemic and because it was prospectively determined using a standardized method in both cohorts. Such findings underscore the consistency between populations and geographical areas of the declining blood testosterone concentrations in older men, a topic of growing interest (1). Our findings in two Australian cities extend previous longitudinal studies to regional centers in another continent and also replicate the findings among men of predominantly European origin (2, 3, 4, 5, 6, 7). Our pooled findings of approximately 1% per annum decline in TT in men over the age of 60 yr, consistently throughout adult life are only marginally lower, compared with these previous longitudinal studies.

One important caveat to these findings is the different assays used by the two cohorts. This makes it difficult to compare directly the absolute concentrations such as the apparently higher blood testosterone and lower SHBG in men from Busselton at baseline (under cross-sectional analysis). For this reason, we used an analytical strategy of examining within-individual (and within method) relative (percentage) annual changes in testosterone or SHBG by linear regression for the longitudinal analyses to reduce potential systematic differences due to assay method. The consistency in the magnitude of the changes in testosterone and SHBG between these two cohorts, particularly after age is controlled for, supports the reliability of these findings and this analytical approach.

Within each cohort, we found that longitudinal exceeded cross-sectional estimates of change substantially (up to 4-fold greater) and additionally steeper falls in calculated free and non-SHBG testosterone, compared with measured TT with either analytical method. These findings are also consistent with other large studies (3, 4).

The significance of the steeper declines in free and non-SHBG bound testosterone is unclear because: 1) the importance of these derived testosterone measures, abstractions calculated as a nonlinear function of TT and SHBG (17), have not been directly tested (19) and 2) these measures overlook putative physiological processes actively mediated by SHBG. Our finding of steeper increases in SHBG with increasing baseline age is therefore intriguing. Increases in SHBG with age have previously been described longitudinally (2, 3); however, our findings additionally show that these rates accelerate with aging, and are independent of obesity, which lowers SHBG.

This study had some limitations. First, as with all observational studies, direct causality cannot be firmly attributed to the relationship between circulating testosterone and age. Second, subjects were older men of mainly European descent living in regional Australia, so generalization to other populations should be cautious. Third, serum was not collected consistently in the morning particularly in the Dubbo cohort, so that circadian rhythms could have contributed to the variability. However, because circadian rhythmicity of blood testosterone concentrations is blunted in older men, diurnal variability is unlikely to have significant impact on our results (20).

We conclude that the age-related decline in serum testosterone and increase in SHBG are comparable across two geographically separated regional Australian communities and also that longitudinal analyses of blood testosterone and SHBG may show up to 4-fold higher changes than cross-sectional analyses, suggesting that longitudinal data are necessary to properly characterize age-related changes. These data extend longitudinal data previously restricted to U.S. and European populations to predominantly Caucasian Australian men.

	Acknowledgments

 
We thank James Modzelewski and Peter Feddema for technical assistance. We thank the Busselton Population Medical Research Foundation for permission to access the stored sera.

	Footnotes

 
Diagnostic Products Corp. (Los Angeles, CA) donated reagents, and funding for laboratory assays was provided by the Royal Perth Hospital Medical Research Foundation and Andrology Australia. P.Y.L. is the recipient of a Neil Hamilton Fairley postdoctoral fellowship (262025) from the National Health and Medical Research Council of Australia. C.M. is the recipient of a medical research fellowship from the Swiss National Science Foundation (81BS-67544) and a research fellowship from the "Margarete und Walter Lichtenstein Stiftung der Universität Basel." The funding sources had no role in the collection, analysis, or interpretation of the data or the decision to submit the manuscript for publication.

Disclosure Statement: The authors have nothing to disclose.

First Published Online June 26, 2007

Abbreviations: ANCOVA, Analysis of covariance; BMI, body mass index; TT, total testosterone.

Received April 17, 2007.

Accepted June 14, 2007.

	References

Top Abstract Introduction Subjects and Methods Results Discussion References

 

1. Liu PY, Swerdloff RS, Veldhuis JD 2004 Clinical review 171: the rationale, efficacy and safety of androgen therapy in older men: future research and current practice recommendations. J Clin Endocrinol Metab 89:4789–4796[Abstract/Free Full Text]
2. Gapstur SM, Gann PH, Kopp P, Colangelo L, Longcope C, Liu K 2002 Serum androgen concentrations in young men: a longitudinal analysis of associations with age, obesity, and race. The CARDIA male hormone study. Cancer Epidemiol Biomarkers Prev 11:1041–1047[Abstract/Free Full Text]
3. Travison TG, Araujo AB, O’Donnell A B, Kupelian V, McKinlay JB 2007 A population-level decline in serum testosterone levels in American men. J Clin Endocrinol Metab 92:196–202[Abstract/Free Full Text]
4. Harman SM, Metter EJ, Tobin JD, Pearson J, Blackman MR 2001 Longitudinal effects of aging on serum total and free testosterone levels in healthy men. Baltimore Longitudinal Study of Aging. J Clin Endocrinol Metab 86:724–731[Abstract/Free Full Text]
5. Morley JE, Kaiser FE, Perry 3rd HM, Patrick P, Morley PM, Stauber PM, Vellas B, Baumgartner RN, Garry PJ 1997 Longitudinal changes in testosterone, luteinizing hormone, and follicle-stimulating hormone in healthy older men. Metabolism 46:410–413[CrossRef][Medline]
6. Zmuda JM, Cauley JA, Kriska A, Glynn NW, Gutai JP, Kuller LH 1997 Longitudinal relation between endogenous testosterone and cardiovascular disease risk factors in middle-aged men. A 13-year follow-up of former Multiple Risk Factor Intervention Trial participants. Am J Epidemiol 146:609–617[Abstract/Free Full Text]
7. T’Sjoen GG, De Vos S, Goemaere S, Van Pottelbergh I, Dierick M, Van Heeringen C, Kaufman JM 2005 Sex steroid level, androgen receptor polymorphism, and depressive symptoms in healthy elderly men. J Am Geriatr Soc 53:636–642[CrossRef][Medline]
8. Gray A, Feldman HA, McKinlay JB, Longcope C 1991 Age, disease, and changing sex hormone levels in middle-aged men: results of the Massachusetts Male Aging Study. J Clin Endocrinol Metab 73:1016–1025[Abstract/Free Full Text]
9. Svartberg J, Midtby M, Bonaa KH, Sundsfjord J, Joakimsen RM, Jorde R 2003 The associations of age, lifestyle factors and chronic disease with testosterone in men: the Tromso Study. Eur J Endocrinol 149:145–152[Abstract]
10. Ferrini RL, Barrett-Connor E 1998 Sex hormones and age: a cross-sectional study of testosterone and estradiol and their bioavailable fractions in community-dwelling men. Am J Epidemiol 147:750–754[Abstract/Free Full Text]
11. Okamura K, Ando F, Shimokata H 2005 Serum total and free testosterone level of Japanese men: a population-based study. Int J Urol 12:810–814[CrossRef][Medline]
12. van den Beld AW, de Jong FH, Grobbee DE, Pols HA, Lamberts SW 2000 Measures of bioavailable serum testosterone and estradiol and their relationships with muscle strength, bone density, and body composition in elderly men. J Clin Endocrinol Metab 85:3276–3282[Abstract/Free Full Text]
13. Belanger A, Candas B, Dupont A, Cusan L, Diamond P, Gomez JL, Labrie F 1994 Changes in serum concentrations of conjugated and unconjugated steroids in 40- to 80-year-old men. J Clin Endocrinol Metab 79:1086–1090[Abstract]
14. Ellison PT, Bribiescas RG, Bentley GR, Campbell BC, Lipson SF, Panter-Brick C, Hill K 2002 Population variation in age-related decline in male salivary testosterone. Hum Reprod 17:3251–3253[Abstract/Free Full Text]
15. Burke V, Knuiman MW, Welborn TA 1998 Overweight and obesity. In: Welborn TA, ed. The Busselton study: mapping population health (cardiovascular and respiratory disease risk factors in Busselton, Australia). North Sydney: Australian Medical Publishing Co.; 142
16. Nguyen T, Sambrook P, Kelly P, Jones G, Lord S, Freund J, Eisman J 1993 Prediction of osteoporotic fractures by postural instability and bone density. BMJ 307:1111–1115[Abstract/Free Full Text]
17. Ly LP, Handelsman DJ 2005 Empirical estimation of free testosterone from testosterone and sex hormone-binding globulin immunoassays. Eur J Endocrinol 152:471–478[Abstract/Free Full Text]
18. Vermeulen A, Verdonck L, Kaufman JM 1999 A critical evaluation of simple methods for the estimation of free testosterone in serum. J Clin Endocrinol Metab 84:3666–3672[Abstract/Free Full Text]
19. Liu PY, Death AK, Handelsman DJ 2003 Androgens and cardiovascular disease. Endocr Rev 24:313–340[Abstract/Free Full Text]
20. Diver MJ, Imtiaz KE, Ahmad AM, Vora JP, Fraser WD 2003 Diurnal rhythms of serum total, free and bioavailable testosterone and of SHBG in middle-aged men compared with those in young men. Clin Endocrinol (Oxf) 58:710–717[CrossRef][Medline]

This article has been cited by other articles:

				D. A. Schaumberg, R. Dana, J. E. Buring, and D. A. Sullivan Prevalence of Dry Eye Disease Among US Men: Estimates From the Physicians' Health Studies Arch Ophthalmol, June 1, 2009; 127(6): 763 - 768. [Abstract] [Full Text] [PDF]

				N. Friedrich, H. Volzke, D. Rosskopf, A. Steveling, A. Krebs, M. Nauck, and H. Wallaschofski Reference Ranges for Serum Dehydroepiandrosterone Sulfate and Testosterone in Adult Men J Androl, November 1, 2008; 29(6): 610 - 617. [Abstract] [Full Text] [PDF]

				B Lapauw, S Goemaere, H Zmierczak, I Van Pottelbergh, A Mahmoud, Y Taes, D De Bacquer, S Vansteelandt, and J M Kaufman The decline of serum testosterone levels in community-dwelling men over 70 years of age: descriptive data and predictors of longitudinal changes Eur. J. Endocrinol., October 1, 2008; 159(4): 459 - 468. [Abstract] [Full Text] [PDF]

				F. C. W. Wu, A. Tajar, S. R. Pye, A. J. Silman, J. D. Finn, T. W. O'Neill, G. Bartfai, F. Casanueva, G. Forti, A. Giwercman, et al. Hypothalamic-Pituitary-Testicular Axis Disruptions in Older Men Are Differentially Linked to Age and Modifiable Risk Factors: The European Male Aging Study J. Clin. Endocrinol. Metab., July 1, 2008; 93(7): 2737 - 2745. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) Submit a related Letter to the Editor Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Copyright Permission Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Liu, P. Y. Articles by Handelsman, D. J. Search for Related Content PubMed PubMed Citation Articles by Liu, P. Y. Articles by Handelsman, D. J. Related Collections Male Endocrinology