Abstract
Background: Aging is characterized by a mild proinflammatory state. In older men, low testosterone levels have been associated with increasing levels of proinflammatory cytokines. It is still unclear whether estradiol (E2), which generally has biological activities complementary to testosterone, affects inflammation.
Methods: We analyzed data obtained from 399 men aged 65–95 yr enrolled in the Invecchiare in Chianti study with complete data on body mass index (BMI), serum E2, testosterone, IL-6, soluble IL-6 receptor, TNF-α, IL-1 receptor antagonist, and C-reactive protein. The relationship between E2 and inflammatory markers was examined using multivariate linear models adjusted for age, BMI, smoking, physical activity, chronic disease, and total testosterone.
Results: In age-adjusted analysis, log (E2) was positively associated with log (IL-6) (r = 0.19; P = 0.047), and the relationship was statistically significant (P = 0.032) after adjustments for age, BMI, smoking, physical activity, chronic disease, and serum testosterone levels. Log (E2) was not significantly associated with log (C-reactive protein), log (soluble IL-6 receptor), or log (TNF-α) in both age-adjusted and fully adjusted analyses.
Conclusions: In older men, E2 is weakly positively associated with IL-6, independent of testosterone and other confounders including BMI.
In a population of older men, total estradiol is positively associated with IL-6, independent of testosterone and other confounders, including body mass index.
Aging in men is associated with a steady decline in serum testosterone levels and with the development of a subclinical inflammatory status (1,2,3). In older men, the decline in testosterone levels is accompanied by a slight decline in estradiol (E2) levels, resulting in an increase of E2/testosterone ratio (4,5,6,7). About 80% of E2 production in men occurs via aromatization of testosterone, especially in the adipose tissue, which is also an important source of inflammatory cytokines (4,8). Some studies suggest a possible relationship between age-related changes in testosterone and E2 and the onset of inflammatory status in older men (9,10,11). We previously reported that testosterone affects inflammation in older men by down-regulating the production of soluble IL-6 receptor (sIL-6r), although the mechanism for this action remains unclear (9). Accordingly, testosterone administration is followed by a marked reduction of inflammatory markers in hypogonadal men (12). Thus, it is reasonable to hypothesize that E2, which generally has biological effects complementary to testosterone, may also affect inflammation. Consistent with this hypothesis, studies have demonstrated that, in adults, E2 is positively associated with C-reactive protein (CRP) levels (13), and, in older men, higher E2 levels are associated with higher risk of stroke, peripheral arterial disease, and diabetes mellitus (14,15,16,17,18). Researchers have suggested that the effect of E2 on inflammation may mediate the relationship between E2 and cardiovascular morbidity (13).
However, to the best of our knowledge, the relationship between total E2 and inflammatory markers has not been previously investigated in a representative population of older men. Using data from the InCHIANTI study, we tested the hypothesis that total E2 is associated with inflammatory markers in older men, independent of testosterone and other confounders.
Subjects and Methods
Study sample
The study population included 417 male residents 65 to 100 yr of age in the area of Chianti, Tuscany, Italy, and enrolled in the Invecchiare in Chianti (InCHIANTI) Study. Of these, 18 were excluded because they were on glucocorticoid (n = 9), antibiotic (n = 6), or androgen (n = 1) treatments or because of recent hospitalization (n = 2), leaving 399 men aged 65 to 95 yr with complete data for analysis. The Italian National Institute of Research and Care on Aging Institutional Review Board ratified the study protocol. Participants consented to participate and to have their blood samples analyzed for scientific purposes (19).
Biological samples
Specimens were collected in a morning fasting state, after at least 15 min of rest; these were conserved in ice for maximum 2 h and delivered to the central laboratory. The serum was stored at −80 C until it was used for analyses.
Laboratory measures and test characteristics
E2 was measured by ultrasensitive RIA (DSL-4800; Diagnostic Systems Laboratories, Webster, TX) with a minimum detectable concentration (MDC) of 2.2 pg/ml and intraassay coefficients of variation (CVs) and means for four different concentrations of 8.9% (5.3 pg/ml), 6.5% (24.9 pg/ml), 7.6% (40.4 pg/ml), and 6.9% (92.6 pg/ml). The interassay CVs and correspondent means were 7.5% (5.3 pg/ml), 9.7% (28.0 pg/ml), 8.0% (42.3 pg /ml), and 12.2% (108.7 pg/ml), respectively. Total testosterone was assayed using commercial radioimmunological kits (Diagnostic Systems Laboratories, Webster, TX). For total testosterone, the MDC was 0.08 ng/ml; intraassay CVs and correspondent means for three different concentrations were 9.6% (0.94 ng/ml), 8.1% (7.01 ng/ml), and 7.8% (19.71 ng/ml), whereas interassay and correspondent means were 8.6% (0.70 ng/ml), 9.1% (5.95 ng/ml), and 8.4% (16.06 ng/ml), respectively (20). Serum IL-6, sIL-6r, TNF-α, IL-1 receptor antagonist (IL-1ra), and CRP were measured in duplicate by high-sensitivity ELISAs (BioSource International, Camarillo, CA). The MDCs were 0.10 pg/ml for IL-6, 8.00 pg/ml for sIL-6r, 0.09 pg/ml for TNF-α, and 4.00 pg/ml for IL-1ra. The CVs were 4.5% for IL-1ra and 7% for IL-6, sIL-6r, and TNF-α. Serum CRP was measured in duplicate by ELISA high-sensitivity test using purified protein and polyclonal anti-CRP antibodies (Calbiochem, San Diego, CA) with standardization according to the World Health Organization First International Reference Standard. The MDC was 0.03 mg/liter, and the CV was 5%.
Study variables
Comorbidities
Comorbidities were ascertained by an experienced clinician according to preestablished criteria that combined information from self-reported physician diagnoses, current pharmacological treatment, medical records, clinical examinations, and blood tests. Comorbidities included in the current analysis were congestive heart failure, diabetes mellitus, hypertension, Parkinson’s disease, peripheral artery disease, cancer, and chronic obstructive pulmonary disease (COPD).
Body size and composition
Weight and height were measured by using standard techniques. Body mass index (BMI) was calculated as weight (in kilograms) divided by the square of height (in meters). The waist circumference was measured at the midpoint between the lower rib margin and the iliac crest (normally umbilical level), and hip circumference was measured at the level of the trochanters. Both were measured to the nearest 0.5 cm with no elastic plastic tape measure to calculate the waist-to-hip ratio.
Health behaviors
Smoking history was determined from self-report and dichotomized in the analysis as “current smoking” vs. “ever smoked” or “never smoked.” Education was assessed as years of school attendance. Physical activity in the year before the interview was coded as: 1) sedentary, completely inactive or light-intensity activity for less than 1 h/wk; 2) light physical activity, light-intensity activity for 2–4 h/wk; and 3) moderate/high physical activity, light activity for at least 5 h/wk or moderate activity for at least 1–2 h/wk. Cognitive function was evaluated using the Mini Mental State Examination (MMSE), and the total score was adjusted for education and age (21). Depressive symptoms were measured by using the 20-item Center for Epidemiological Studies Depression Scale (CES-D). Participants with a CES-D score of at least 16 were considered to be depressed (22).
Daily total energy (kilocalories) and alcohol (grams) intake were estimated by the European Prospective Investigation into Cancer and Nutrition food frequency questionnaire (23).
Disability was evaluated by asking participants whether they were able to perform several basic activities of daily living (ADL) without the help of another person. Participants who reported need for help in one or more activities were considered as disabled (19).
Statistical analysis
Variables are reported as mean ± sd, medians and interquartile ranges, or number and percentages. Because of skewed distributions, log-transformed values for E2, IL-6, sIL-6r, TNF-α, CRP and IL-1ra, CES-D score, MMSE score, education, and daily alcohol consumption were used in regression analyses and back-transformed for data presentation when necessary. Age-adjusted IL-6 was explored by a scatter plot with superimposed linear regression line. Factors statistically correlated with log (IL-6) and log (E2) were identified using age-adjusted Pearson partial correlation coefficient. Those variables correlated with either log (IL-6) or E2 were considered potential confounders, and their intercorrelation was examined using an age-adjusted Pearson correlation matrix. Because none of the r-values in the correlation matrix was greater than 0.3, they were all introduced as potential confounders in the regression model that examined the relationship between E2 and IL-6. To test the hypothesis that the relationship between estrogens and inflammation simply reflects the effect of testosterone (9), we introduced testosterone as a further potential confounder in the fully adjusted regression model testing the association between E2 and inflammatory markers. This analysis was limited to participants free of ADL disability (n = 382; 95.7%).
Analyses were performed by the SAS statistical package, version 9.1 (SAS Institute Inc., Cary, NC).
Results
The characteristics of the study population are summarized in Table 1. The median level of E2 was 12.95 pg/ml (interquartile range, 9.63–16.55) and of IL-6 was 1.56 pg/ml (interquartile range, 0.92–2.54). After adjusting for age, log (E2) was positively associated with log (IL-6) (r = 0.19; P = 0.047) (Table 2 and Fig. 1). Log (IL-6) was also higher in participants who were current smokers (P = 0.0002) and those affected by chronic heart failure (CHF) (P = 0.01), peripheral artery disease (P = 0.03), and hip osteoarthritis (P = 0.07). Log (IL-6) was negatively associated with level of physical activity (P = 0.005) and nonsignificantly with BMI (P = 0.08) (Table 2).
Table 1.
Characteristics of the study population (n = 399)
| Characteristic | |
|---|---|
| Estradiol (pg/ml) | 12.95 [9.63–16.55] |
| Inflammatory markers | |
| IL-6 (pg/ml) | 1.56 [0.92–2.54] |
| sIL-6r (ng/liter) | 91.32 [67.12–129.89] |
| IL-1ra (pg/ml) | 130.08 [95.23–183.02] |
| TNF-α (pg/ml) | 0.00 [0.00–1.50] |
| CRP (mg/liter) | 2.97 [1.30–6.18] |
| Age (yr) | 74.8 ± 7.0 |
| BMI (kg/m2) | 27.2 ± 3.1 |
| Waist-to-hip ratio | 0.96 [0.93–0.99] |
| Total testosterone (ng/ml) | 4.24 ± 1.32 |
| Disease | |
| Cancer | 18 (4.5) |
| Diabetes | 57 (14.3) |
| Hip osteoarthritis | 31 (7.8) |
| Knee osteoarthritis | 62 (15.5) |
| Peripheral artery disease | 37 (9.2) |
| Angina pectoris | 9 (2.2) |
| Parkinson disease | 9 (2.2) |
| COPD | 15 (3.8) |
| CHF | 82 (20.5) |
| Depressive symptoms (CES-D score) | 8 [5–13] |
| Cognitive status (MMSE score) | 26 [24–28] |
| Social and behavioral factors | |
| Log (alcohol intake) | 20.0 [10–30] |
| Current smoking | 86 (21.5) |
| Education | 5 [5–7] |
| Physical activity | |
| Sedentary | 57 (14.2) |
| Light | 309 (77.6) |
| Moderate/high | 33 (8.2) |
Values are expressed as median [interquartile range], means ± sd, or number (%).
Table 2.
Age-adjusted Pearson partial correlation coefficients of log (IL-6) and log (E2) and other factors
| Characteristic (n = 399) | Log (IL-6)
|
|
|---|---|---|
| r | P value | |
| Log (E2) | 0.19 | 0.047 |
| Log (inflammatory markers) | ||
| sIL-6r | 0.11 | 0.15 |
| IL-1ra | 0.37 | <0.0001 |
| TNF-α | 0.04 | 0.61 |
| CRP | 0.58 | <0.0001 |
| BMI | −0.13 | 0.08 |
| Log (waist-to-hip ratio) | −0.08 | 0.63 |
| Disease | ||
| Cancer | 0.96 | 0.20 |
| Diabetes | −0.02 | 0.68 |
| Hip osteoarthritis | 0.09 | 0.07 |
| Knee osteoarthritis | 0.06 | 0.22 |
| Peripheral artery disease | 0.11 | 0.03 |
| Parkinson disease | −0.05 | 0.46 |
| COPD | 0.02 | 0.73 |
| CHF | 0.19 | 0.01 |
| Log (CES-D score) | 0.05 | 0.36 |
| Log (MMSE score) | 0.04 | 0.40 |
| Social and behavioral factors | ||
| Log (alcohol intake) | 0.03 | 0.59 |
| Current smoking | 0.19 | 0.0002 |
| Physical activity | −0.14 | 0.005 |
| Log (education) | −0.047 | 0.36 |
Figure 1.
Scatterplot of the relationship between E2 and IL-6. IL-6 values are age-adjusted using a residual method, and the relationship is statistically significant (P = 0.047).
In a multivariate regression model adjusted for age, BMI, physical activity, current smoking, hip osteoarthritis, peripheral artery disease, and CHF, the relationship between log (E2) and log (IL-6) remained significant (P = 0.047; R2 = 0.14) (Table 3). The relationship between log (E2) and log (IL-6) was unchanged after further adjustment for total testosterone (P = 0.032; R2 = 0.15) (Table 3).
Table 3.
Relationship of E2 with IL-6, sIL-6r, TNF-α, IL-1ra, and CRP
| Log (E2) (n = 399)
|
||||
|---|---|---|---|---|
| βα | se | P | ||
| Model 1 (fully adjusted) | ||||
| Log IL-6 | [R2 = 0.14] | 0.02 | 0.007 | 0.047 |
| Log sIL-6r | [R2 = 0.04] | 0.001 | 0.004 | 0.84 |
| Log TNF-α | [R2 = 0.03] | −0.012 | 0.004 | 0.48 |
| Log IL-1ra | [R2 = 0.03] | 0.02 | 0.005 | 0.64 |
| Log CRP | [R2 = 0.09] | 0.002 | 0.01 | 0.87 |
| Model 2 (model 1 + total testosterone) | ||||
| Log IL-6 | [R2 = 0.15] | 0.02 | 0.008 | 0.032 |
| Log sIL-6r | [R2 = 0.07] | 0.006 | 0.005 | 0.23 |
| Log TNF-α | [R2 = 0.04] | −0.021 | 0.02 | 0.27 |
| Log IL-1ra | [R2 = 0.03] | 0.006 | 0.005 | 0.31 |
| Log CRP | [R2 = 0.09] | 0.001 | 0.01 | 0.02 |
Each line refers to the results of a separate model adjusted for indicated covariates. Values were adjusted for age, BMI, physical activity, smoking, hip osteoarthritis, CHF, and peripheral artery disease (model 1).
In the fully adjusted regression model, log (E2) was not associated with log (sIL-6r), log (TNF-α), log (IL-1ra), or log (CRP) (Table 3). Excluding participants with ADL disability, the results remained substantially unchanged.
Discussion
In this representative sample of older men, we demonstrated that total E2 levels are weakly but significantly associated with IL-6 after adjusting for age and multiple potential confounders, including total testosterone. E2 was not associated with other inflammatory markers. Contrary to our hypothesis, the association between E2 and IL-6 was strengthened after adjusting for testosterone, suggesting that E2 does not mediate the relationship between testosterone and sIL-6r previously reported in the InCHIANTI population (9).
E2, the major biologically active estrogen, is a metabolite of testosterone, and in adult men about 80% of it originates from aromatization of testosterone, especially in adipose tissue, whereas less than 20% of circulating E2 is secreted by testis (4). This explains why older men tend to have higher E2 levels than postmenopausal women, suggesting that higher testosterone levels and BMI, a proxy measure of adipose tissue, may account for E2 concentration in older men (24). However, when we adjusted for total testosterone and BMI, the relationship between E2 and IL-6 did not change (Table 3).
E2 may affect or reflect insulin and insulin resistance. At least in this study we found no significant relationship between E2 and fasting insulin (β coefficient ± se = −0.002 ± 0.003; P = 0.64) as well between E2 and insulin resistance (homeostasis model assessment) (β coefficient ± se = 0.007 ± 0.005; P = 0.20). The correlation between E2 and IL-6 levels in this representative sample of older men, although moderate, is a novel finding. Previous studies testing the relationship between E2 and IL-6 are limited to conditions of estrogen deficiency such as hypogonadism induced by GnRH agonists (25). Khosla et al. (25) showed that the induction of hypogonadism with low testosterone and E2 levels resulted in a significant increase in IL-6 and sIL-6r concentration in adults. However, whether the suppression of testosterone rather than the decrease in E2 levels was the main cause of inflammation remained unclear.
Studies conducted in selective models of estrogen deficiency, namely the congenital aromatase deficiency, showed an increased prevalence of insulin resistance that may be caused by inflammation and cardiovascular disease but did not provide information on IL-6 and other inflammatory markers (26). By contrast, other studies have shown that E2 and, in particular, the E2/testosterone ratio are positively associated with adipokines other than IL-6, such as leptin, and inversely related to adiponectin in older adults, suggesting a potential link between E2 and unfavorable metabolic status in older persons (27).
We did not observe any significant association between endogenous E2 levels and CRP. This is consistent with the finding of a previous observational study conducted in 715 healthy young and middle-aged men (13) and with the lack of effect on CRP level after E2 transdermal treatment in men with prostate cancer (28). Interestingly, Nakhai Pour et al. (13) showed that E2 levels were positively associated with CRP, but the association was no longer statistically significant after adjustment for intraabdominal fat, suggesting that adipose tissue, which is an important source of aromatase and inflammatory cytokines, may mediate the relationship between E2 and inflammation (16). However, when we adjusted for many confounders including BMI and waist-to-hip ratio in our study, the relationship between E2 and IL-6 was still significant. Even when we analyzed the effect of waist circumference, which is probably a more accurate measure of central fat than waist-to-hip ratio, the relationship between E2 and IL-6 remained statistically unchanged (β ± se = 0.02 ± 0.01; P = 0.01). Beyond IL-6, in this study, we tested the relationship between E2 and sIL-6r, a soluble form of IL-6 receptor that enhances the IL-6 activity by making cells that have a glycoprotein 130 membrane domain sensitive to IL-6 signaling. Because we found no relationship between E2 and sIL-6r, we suggest that the relationship between testosterone and sIL-6r, which has been previously reported, is not mediated by E2 (9). The weak but significant association of E2 with IL-6 compared with the other proinflammatory cytokines is puzzling and remained unexplained, but it is consistent with other studies showing that the mild proinflammatory state of aging is best captured by IL-6 and not other inflammatory markers, possibly because the sensitivity and reliability of the IL-6 assay currently available are superior to those for the assays of other cytokines (3,29).
Our study has limitations. We did not measure free E2, and therefore we have only a rough measure of estrogen signaling. Future studies should measure estrone, androstenedione, and E2 receptor expression to better characterize the whole estrogen signaling. The measure of IL-6 in the peripheral blood is by its own nature an approximation. Finally, given the cross-sectional nature of this study, and because of the moderate correlation between E2 and IL-6, the findings of this study can only suggest the existence of a connection between E2 and IL-6 but does not shed light on the nature of this connection and cannot prove any causal relationship. These limitations are offset by important strengths; first, we used measured E2 and cytokine levels obtained with RIA and ultrasensitive ELISA, which are highly sensitive assays. Second, by excluding participants who were on antibiotics and steroids and those recently hospitalized, we selected a healthier population. Third, the wealth of information on markers, hormones, and relevant covariates collected in the InCHIANTI study is almost a unique feature of this project.
In conclusion, in a representative sample of older men, total E2 levels were not affected by age. E2 was weakly and positively associated with IL-6, but not with other inflammatory markers, including sIL-6r. Replication of our findings in longitudinal studies is required to elucidate the direction of the relationship between E2 and IL-6, as well as the physiological significance of these cross-sectional observations.
Footnotes
The InCHIANTI Study was supported as a “targeted project” (ICS 110.1/RS97.71) by the Italian Ministry of Health and in part by the U.S. National Institute on Aging (contracts N01-AG-916413 and N01-AG-821336) and by the Intramural Research Program of the U.S. National Institute on Aging (contracts 263 MD 9164 13 and 263 MD 821336). None of the sponsoring institutions interfered with the collection, analysis, presentation, or interpretation of the data reported here.
Conflict of Interests: The authors declare that they have no conflict of interest to disclose concerning this manuscript.
First Published Online December 2, 2008
Abbreviations: BMI, Body mass index; CES-D, Center for Epidemiological Studies Depression Scale; CHF, chronic heart failure; COPD, chronic obstructive pulmonary disease; CRP, C-reactive protein; CV, coefficient of variation; E2, estradiol; IL-1ra, IL-1 receptor antagonist; MDC, minimum detectable concentration; MMSE, Mini Mental State Examination; sIL-6r, soluble IL-6 receptor.
References
- 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 [DOI] [PubMed] [Google Scholar]
- 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 [DOI] [PubMed] [Google Scholar]
- Ferrucci L, Corsi A, Lauretani F, Bandinelli S, Bartali B, Taub DD, Guralnik JM, Longo DL 2005 The origins of age-related proinflammatory state. Blood 105:2294–2299 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Vermeulen A, Kaufman JM, Goemaere S, van Pottelberg I 2002 Estradiol in elderly men. Aging Male 5:98–102 [PubMed] [Google Scholar]
- 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 [DOI] [PubMed] [Google Scholar]
- 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 [DOI] [PubMed] [Google Scholar]
- Orwoll E, Lambert LC, Marshall LM, Phipps K, Blank J, Barrett-Connor E, Cauley J, Ensrud K, Cummings S 2006 Testosterone and estradiol among older men. J Clin Endocrinol Metab 91:1336–1344 [DOI] [PubMed] [Google Scholar]
- Longcope C, Pratt JH, Schneider SH, Fineberg SE 1978 Aromatization of androgens by muscle and adipose tissue in vivo. J Clin Endocrinol Metab 46:146–152 [DOI] [PubMed] [Google Scholar]
- Maggio M, Basaria S, Ble A, Lauretani F, Bandinelli S, Ceda GP, Valenti G, Ling SM, Ferrucci L 2006 Correlation between testosterone and the inflammatory marker soluble interleukin-6 receptor in older men. J Clin Endocrinol Metab 91:345–347 [DOI] [PubMed] [Google Scholar]
- Maggio M, Cappola AR, Ceda GP, Basaria S, Chia CW, Valenti G, Ferrucci L 2005 The hormonal pathway to frailty in older men. J Endocrinol Invest 28 (11 Suppl Proceedings):15–19 [PubMed] [Google Scholar]
- Ferrucci L, Harris TB, Guralnik JM, Tracy RP, Corti MC, Cohen HJ, Penninx B, Pahor M, Wallace R, Havlik RJ 1999 Serum IL-6 level and the development of disability in older persons. J Am Geriatr Soc 47:639–646 [DOI] [PubMed] [Google Scholar]
- Malkin CJ, Pugh PJ, Jones RD, Kapoor D, Channer KS, Jones TH 2004 The effect of testosterone replacement on endogenous inflammatory cytokines and lipid profiles in hypogonadal men. J Clin Endocrinol Metab 89:3313–3318 [DOI] [PubMed] [Google Scholar]
- Nakhai Pour HR, Grobbee DE, Muller M, van der Schouw YT 2007 Association of endogenous sex hormone with C-reactive protein levels in middle-aged and elderly men. Clin Endocrinol (Oxf) 66:394–398 [DOI] [PubMed] [Google Scholar]
- Phillips GB, Castelli WP, Abbott RD, McNamara PM 1983 Association of hyperestrogenemia and coronary heart disease in men in the Framingham Cohort. Am J Med 74:863–869 [DOI] [PubMed] [Google Scholar]
- Phillips GB, Pinkernell BH, Jing TY 1996 The association of hyperestrogenemia with coronary thrombosis in men. Arterioscler Thromb Vasc Biol 16:1383–1387 [DOI] [PubMed] [Google Scholar]
- Barrett-Connor E, Khaw KT 1988 Endogenous sex hormones and cardiovascular disease in men: a prospective population based study. Circulation 78:539–545 [DOI] [PubMed] [Google Scholar]
- Tivesten A, Mellstrom D, Jutberger H, Fagerberg B, Lernfelt B, Orwoll E, Karlsson MK, Ljunggren O, Ohlsson C 2007 Low serum testosterone and high serum estradiol associate with lower extremity peripheral arterial disease in elderly men. The MrOS Study in Sweden. J Am Coll Cardiol 50:1070–1076 [DOI] [PubMed] [Google Scholar]
- Abbott RD, Launer LJ, Rodriguez BL, Ross GW, Wilson PW, Masaki KH, Strozyk D, Curb JD, Yano K, Popper JS, Petrovitch H 2007 Serum estradiol and risk of stroke in elderly men. Neurology 68:563–568 [DOI] [PubMed] [Google Scholar]
- Ferrucci L, Bandinelli S, Benvenuti E, Di Iorio A, Macchi C, Harris TB, Guralnik JM 2000 Subsystems contributing to the decline in ability to walk: bridging the gap between epidemiology and geriatric practice in the InCHIANTI study. J Am Geriatr Soc 48:1618–1625 [DOI] [PubMed] [Google Scholar]
- Valenti G, Denti L, Maggio M, Ceda G, Volpato S, Bandinelli S, Ceresini G, Cappola A, Guralnik JM, Ferrucci L 2004 Effect of DHEA-S on skeletal muscle over the life span: the InCHIANTI study. J Gerontol A Biol Sci Med Sci 59:466–472 [DOI] [PubMed] [Google Scholar]
- Folstein MF, Folstein SE, McHugh PR 1975 Mini-mental state. A practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res 12:189–198 [DOI] [PubMed] [Google Scholar]
- Myers JK, Weissman MM 1980 Use of a self-report symptom scale to detect depression in a community sample. Am J Psychiatry 137:1081–1084 [DOI] [PubMed] [Google Scholar]
- Pisani P, Faggiano F, Krogh V, Palli D, Vineis P, Berrino F 1997 Relative validity and reproducibility of a food frequency dietary questionnaire for use in the Italian EPIC centres. Int J Epidemiol 26(Suppl 1):S152–S160 [DOI] [PubMed] [Google Scholar]
- Jankowska EW, Medras M, Rogucka EA 2000 Body mass index, waist/hip ratio and androgen-estrogen activity in younger versus older Polish men. Aging Male 3:177–184 [DOI] [PubMed] [Google Scholar]
- Khosla S, Atkinson EJ, Dunstan CR, O'Fallon WM 2002 Effect of estrogen versus testosterone on circulating osteoprotegerin and other cytokine levels in normal elderly men. J Clin Endocrinol Metab 87:1550–1554 [DOI] [PubMed] [Google Scholar]
- Rochira V, Balestrieri A, Madeo B, Spaggiari A, Carani C 2002 Congenital estrogen deficiency in men: a new syndrome with different phenotypes; clinical and therapeutic implications in men. Mol Cell Endocrinol 193:19–28 [DOI] [PubMed] [Google Scholar]
- Laughlin GA, Barrett-Connor E, May S 2006 Sex–specific association of the androgen to oestrogen ratio with adipocytokine levels in older adults: the Rancho Bernardo Study. Clin Endocrinol (Oxf) 65:506–513 [DOI] [PubMed] [Google Scholar]
- Purnell JQ, Bland LB, Garzotto M, Lemmon D, Wersinger EM, Ryan CW, Brunzell JD, Beer TM 2006 Effects of transdermal estrogen on levels of lipids, lipase activity, and inflammatory markers in men with prostate cancer. J Lipid Res 47:349–355 [DOI] [PubMed] [Google Scholar]
- Maggio M, Guralnik JM, Longo DL, Ferrucci L 2006 Interleukin-6 in aging and chronic disease: a magnificent pathway. J Gerontol A Biol Sci Med Sci 61:575–584 [DOI] [PMC free article] [PubMed] [Google Scholar]