J Endocrinol Invest (2014) 37:135–141 DOI 10.1007/s40618-013-0009-x
ORIGINAL ARTICLE
Hypogonadism in aged hospitalized male patients: prevalence and clinical outcome P. Iglesias • F. Prado • M. C. Macı´as • M. T. Guerrero • A. Mun˜oz • E. Ridruejo • P. Tajada • C. Garcı´a-Are´valo • J. J. Dı´ez
Received: 3 September 2013 / Accepted: 17 November 2013 / Published online: 9 January 2014 Ó Italian Society of Endocrinology (SIE) 2013
Abstract Objective Male hypogonadism is common in the elderly and has been associated with increased risk of mortality. Our objective has been to assess the prevalence of primary and central hypogonadism in elderly male patients admitted to the hospital because of acute illness. We also evaluated the relationships between gonadal dysfunction and in-hospital mortality. Patients and methods 150 patients, aged C65 years, admitted during 2010 and 2011 in our geriatric unit, were studied. Serum concentrations total, bioavailable and free testosterone, as well as of follicle-stimulating hormone and luteinizing hormone were quantified in every patient. Hypogonadism was defined by the presence of serum testosterone levels lower than 200 ng/dl. Results Hypogonadism was found in 80 patients (53.3 %). Serum gonadotropin concentrations were elevated in 43.7 % of these patients, whereas 41.3 % of hypogonadic patients showed normal and 15 % low gonadotropin concentrations. Respiratory tract infection and congestive heart failure were the main causes of hospitalization in hypogonadal men, whereas acute cerebrovascular disease was the main reason for admission in eugonadal patients. Of the 13 patients who died during hospitalization, 12 were hypogonadic. Patients
P. Iglesias (&) J. J. Dı´ez Department of Endocrinology, Hospital Ramo´n y Cajal, Ctra. de Colmenar, Km 9,100, 28034 Madrid, Spain e-mail: [email protected] F. Prado M. C. Macı´as M. T. Guerrero A. Mun˜oz E. Ridruejo Department of Geriatrics, Hospital General, Segovia, Spain P. Tajada C. Garcı´a-Are´valo Department of Biochemistry, Hospital General, Segovia, Spain
who died showed significantly lower serum levels of total, free and bioavailable testosterone than those found in patients who survived. Conclusion Our results show that about half of patients admitted for acute illness have hypogonadism, mainly of non-hypergonadotropic type. Gonadal hypofunction is significantly related with in-hospital mortality. A low value of serum testosterone may be a predictor for mortality in elderly male patients. Keywords Aged male patients Elderly Testosterone Hypogonadism Clinical outcome
Introduction Epidemiological surveys have shown that the prevalence of hypogonadism ranges from 6 to 9.5 % in communitydwelling men aged 40–70 years, and rises to 15–30 % in diabetic or obese men [1]. Due to the fact that gonadal function declines gradually with age in men, testosterone (T) deficiency is even more prevalent in the elderly ranging from 30 to 50 % [2–9]. An association between low T levels and morbidity and mortality in elderly men has been reported. Several chronic disease including diabetes mellitus, chronic obstructive pulmonary disease, inflammatory arthritis, kidney failure and HIV infection, obesity, metabolic syndrome and hemochromatosis are well-known risk factors for hypogonadism in the elderly [7, 10, 11]. On the other hand, a low serum T concentration has been considered as a cardiovascular risk factor in the elderly [12, 13]. Lastly, several large studies, although not all [14–16], have found a relationship between low T levels and cardiovascular and overall mortality in elderly men [12, 13, 17–19].
123
136
To the best of our knowledge, the prevalence and clinical significance of gonadal dysfunction in elderly male patients admitted for acute illness has not clearly been evaluated. We performed a cross-sectional study in a cohort of elderly male patients admitted to our hospital for acute disease. Our aim has been to assess the prevalence of hypogonadism in this population, as well as to study the relationship between gonadal derangements and morbidity and in-hospital mortality in these patients.
Patients and methods
J Endocrinol Invest (2014) 37:135–141
Diagnostics Inc., Tarrytown, NY, USA). The sensitivity (S) of the assay for these hormones was 0.3 mU/ml, 0.07 mU/ml, 0.3 ng/ml, 10 pg/ml, and 10 ng/dl, respectively. Normal ranges (NRs) for adult males were: FSH 1.4–18.1 mU/ml; LH 1.5–9.3 mU/ml; PRL 2.1–17.7 ng/ml; E2 11.6–41.2 pg/ml; and T 240–820 ng/dl. Other serum hormones analyzed by a chemiluminescent immunometric assay (Immulite 2,000 analyzer, Siemens Medical Solutions Diagnostics, Llanberis, Gwynedd, UK) were D-4 (NR 0.6–3.1 ng/ml; S 0.3 ng/ml), and DHEA-S (NR 80–560 mcg/dl; S 3 mcg/dl). Finally, SHBG was quantified by the same procedure and NR and S were 10–57 and 0.02 nmol/l, respectively.
Patients Criteria for diagnosis A cross-sectional study including all male patients older than 65 years that were hospitalized during two consecutive years (2010 and 2011) was carried out. All patients admitted to our geriatric unit were included in the study regardless of the cause of hospitalization. A detailed clinical history and complete physical examination were performed in every patient. Functional status at entry was evaluated using the modified version of Barthel index [20]. This index evaluates ten basic activities of everyday life with a score that ranges from 0 (maximal dependence) to 100 (maximal independence). Previous comorbidity and main cause for hospitalization were also registered and related to gonadal dysfunction. All deaths during hospitalization were registered. This study was approved by our local ethical committee and informed consent was given by subjects prior to participation in the study.
Patients were classified into eugonadal and hypogonadal (hyper-, normo- or hypogonadotropic) depending on the values of serum T and gonadotropins (FSH and LH). In this elderly population, patients were considered eugonadal when serum T concentrations were C200 ng/dl regardless of the gonadotropins values. The diagnosis of primary hypogonadism (hypergonadotropic) was established when patients simultaneously showed serum T concentrations \200 ng/dl and FSH and LH concentrations exceeding the upper limit of the normal reference range for these hormones ([18.1 mU/ml for FSH and [9.3 mU/ml for LH). Central or secondary hypogonadism (normo or hypogonadotropic) was considered when reduced serum T concentrations (\200 ng/dl) and normal or decreased gonadotropins levels (B18.1 mU/ml for FSH and B9.3 mU/ml for LH) were found.
Hormonal gonadal study
Statistical analysis
Gonadal function was assessed by measuring serum concentrations of gonadotropins (follicle-stimulating hormone, FSH and luteinizing hormone, LH) and total testosterone (T). Bioavailable (BioT) and free testosterone (FT) were calculated based on the values of T, sex hormone binding globulin (SHBG), and albumin, as previously described [21]. Other serum hormones analyzed were estradiol (E2), prolactin (PRL), delta-4-androstenedione (D-4), and dehydroepiandrosterone sulfate (DHEA-S).
For quantitative variables, results are expressed as mean ± SD for normally distributed data, and as median (range) for nonparametric data. Adjustment to normal distribution was tested by the Kolmogorov test. Categorical variables are described as percentages. For comparisons of means between two groups of subjects the Student t test was used for normally distributed data, and the Mann– Whitney test was employed for nonparametric data. For ratio comparisons the Chi square test and Fisher’s exact test were used. Correlations between quantitative variables were assessed using Spearman’s correlation analysis. Multivariate regression analysis between serum testosterone (dependent variable) and clinical and analytical variables was also evaluated. Models of logistic regression analysis were used to assess hospital mortality as a function of the different hormones analyzed. Differences were considered significant when p \ 0.05.
Hormone assays Fasting samples of venous blood were obtained from an antecubital vein between 08:30 and 09:00 h in the first day of admission for quantification. Serum FSH, LH, PRL, E2, and T were measured by a chemiluminiscence immunoassay (ADVIA Centaur XP analyzer, Siemens Healthcare
123
J Endocrinol Invest (2014) 37:135–141
137
Table 1 Clinical and analytical data of the 150 studied patients at hospital admission Mean ± SD or median (interquartile range)
Range (min/ max)
Age (years)
86.0 ± 4.8
68–99
Barthel index at admission
80 (50–90)
0–100
BMI (kg/m2)
25.7 ± 4.0
13–46
SBP (mmHg)
128 ± 24
60–193
DBP (mmHg)
70 ± 14
30–113
Clinical data
Analytical data FSH (mU/ml)
8.2 (4.9–14.0)
0.7–81.3
LH (mU/ml)
9.0 (5.1–14.7)
0.06–80.2
PRL (ng/ml)
9.5 (7.0–13.9)
1.1–96.9
Estradiol (pg/ml)
19.4 (6.9–32.0)
5.9–222.0
Testosterone (ng/dl)
190 (90–330)
10–970
SHBG (nmol/l)
68 (34–67)
13–146
Albumin (g/dl) Free testosterone (ng/dl)
3.4 ± 0.5 3.2 (1.4–5.3)
1.6–4.6 0.1–16.0
Free testosterone (%)
1.7 ± 0.7
0.6–3.5
Bioavailable testosterone (ng/ dl)
58.5 (28.4–105.7)
2.1–363
Table 2 Distribution of the patients according to serum concentrations of gonadotropins and gonadal status Hypogonadism (Serum testosterone \200 ng/dl) Yes
Total
No
FSH Low
3 (3.8)
0
3 (2.0)
Normal
60 (75.0)
58 (82.9)
118 (78.7)
High
17 (21.2)
12 (17.1)
29 (19.3)
Low
12 (15.0)
0
12 (8.0)
Normal
33 (41.2)
36 (51.4)
69 (46.0)
High
35 (43.8)
34 (48.6)
69 (46.0)
LH**
FSH and LH* Both low
3 (3.7)
0
3 (2.0)
Both normal
33 (41.3)
35 (50.0)
68 (45.3)
Both high
17 (21.3)
11 (15.7)
28 (18.7)
FSH normal/LH low
9 (11.3)
0
9 (6.0)
FSH normal/LH high LH normal/FSH low
18 (22.4) 0
23 (32.9) 0
41 (27.3) 0
0
1 (1.4)
1 (0.7)
80 (53.3)
70 (46.7)
150
LH normal/FSH high Total
Bioavailable testosterone (%)
31.9 ± 10.6
10.4–63.1
Figures indicate the number of patients (%)
Delta 4-androstenedione (ng/ ml)
1.0 (0.6–1.4)
0.02–4.6
Hypogonadism vs. normal testosterone
DHEA-S (mcg/dl)
220 (149–455)
28–2030
* p = 0.014
BMI body mass index, SBP systolic blood pressure, DBP diastolic blood pressure, LH luteinizing hormone, FSH follicle-stimulating hormone, PRL prolactin, SHBG sex hormone binding globulin, DHEA-S dehydroepiandrosterone sulfate
Results Clinical features and analytical data A group of 150 male patients (mean age ± SD, 86.0 ± 4.8 years) was studied. The main causes of admission at the hospital were: congestive heart failure (28 patients, 18.7 %), acute cerebrovascular disease (21 patients, 14.0 %), respiratory tract infection (28 patients, 18.7 %), acute digestive hemorrhage (13 patients, 8.7 %), coronary heart disease (8 patients, 5.3 %), exacerbation of chronic obstructive pulmonary disease (7 patients, 4.7 %), cancer (4 patients, 2.7 %), anemia (3 patients, 2.0 %), and other (38 patients, 25.2 %). Clinical and hormonal data of the studied patients are summarized in Table 1. Prevalence of gonadal dysfunction Hypogonadism was present in about half of the patients (80 patients, 53.3 %). Among them, 43.7 % were
** p = 0.003
hypergonadotropic, 41.3 % normogonadotropic and 15 % hypogonadotropic. Seventy (46.7 %) patients were eugonadal, half of them (35 patients, 50 %) normogonadotropic and the other half hypergonadotropic. We found a statistically significant (p = 0.003) difference in the proportion of hypo-, normo- and hypergonadotropic patients between patients with and without hypogonadism, mainly due to the presence of a greater percentage of hypogonadotropic patients in the hypogonadal group. Table 2 shows the distribution of the patients according to the level of serum gonadotropins and the presence or absence of hypogonadism. Regression analysis In the bivariate correlation analysis, serum T concentration was positively correlated with Barthel index (q = 0.280, p = 0.001), systolic blood pressure (SBP) (q = 0.212, p = 0.009), albumin (q = 0.334, p \ 0.001), SHBG (q = 0.265, p = 0.001), E2 (q = 0.405, p \ 0.001), and D-4 (q = 0.318, p \ 0.001). On the contrary, T was negatively correlated only with age (q = -0.301, p \ 0.001). Multivariate regression analysis showed a significant correlation between serum T concentration (dependent
123
138
J Endocrinol Invest (2014) 37:135–141
variable) and E2 (b = 0.187, p = 0.018), albumin (b = 0.234, p = 0.002), and SHBG (b = 0.270, p = 0.000). Gonadal function and associated diseases There was an association between hypogonadism and personal history of heart disease (heart failure, angina pectoris and/or acute myocardial infarction) (p \ 0.032), cancer (p = 0.004), respiratory pathology (p = 0.016), and renal insufficiency (p = 0.013). However, we could not find any association between hypogonadal status and hypertension, diabetes, hyperlipidemia, anemia, and prostate pathology. A significant relationship between gonadal status and the main cause of hospitalization was also found (p = 0.034). In this regard, the three main causes of admission for hypogonadic patients were respiratory tract infection (23.7 %), congestive heart failure (22.5 %), and acute cerebrovascular disease (6.2 %); however, in the eugonadal group these percentages were 12.8, 14.3 and 22.8 %, respectively. Moreover, patients admitted by respiratory tract infection and acute cerebrovascular disease showed the lowest and the highest values of T concentration [103 ng/dl (44–236 ng/dl) vs. 466 ng/dl (195–610 ng/dl), p \ 0.05]. Gonadal function and mortality Thirteen patients (8.7 %) died during hospitalization. Hypogonadism was associated to mortality during hospital stay. In fact, most of the patients (12 out of 13) who died were hypogonadic compared with 49.6 % of the patients who survived (p = 0.003) (Fig. 1). Patients who died showed significantly lower serum T (p \ 0.01), FT (p \ 0.01), and BioT (p \ 0.01) than survivors (Table 3). In-hospital mortality was associated with a significantly lower serum T concentrations both in very old patients (C85 year) (p = 0.013) and in those with a body mass index (BMI) \25 kg/m2 (p = 0.005) but not in patients \85 year and in those with a BMI C25 kg/m2. Three models of logistic regression analysis performed to study the influence of assessed hormones and clinical parameters (age and BMI) on fatal outcome during hospitalization showed that serum levels of T and D-4 were independent predictors of mortality in hospitalized elderly men population (Table 4). However, after adjusting for age and BMI, only T, but not D-4, showed a significant relationship with hospital mortality [OR 0.993 (0.987–0.999), P = 0.02]. Lastly, according to a model of logistic regression analysis, the likelihood of death during hospitalization in elderly men increased at lower serum T concentrations, ranging between 0.07 and 0.23 (Fig. 2).
123
Fig. 1 Percent distribution of the patients according to gonadal status at entry and hospital mortality Table 3 Distribution of the patients according to vital status at discharge Hospital mortality Yes (n = 13, 8.7 %)
No (n = 137, 91.3 %)
FSH (mU/ml)
6.1 (3.0–20.2)
8.3 (5.1–14.0)
LH (mU/ml)
6.2 (5.0–27.6)
9.1 (5.1–14.2)
PRL (ng/ml)
11.9 (9.4–25.0)
9.3 (6.9–13.9)
Estradiol (pg/ml)
22.5 (11.0–32.2)
19.0 (6.9–32.1)
Testosterone (ng/dl)
89 (39–158)**
201 (104–369)
Free testosterone (ng/dl)
1.4 (0.5–2.3)**
3.6 (1.7–5.6)
Free testosterone (%)
1.74 ± 0.56
1.68 ± 0.57
Bioavailable testosterone (ng/dl) Bioavailable testosterone (%)
28.8 (5.8–45.1)**
68.8 (30.1–110.0)
30.2 ± 11.8
32.1 ± 10.5
Delta 4-androstenedione (ng/ml)
1.4 (0.6–2.0)
1.0 (0.6–1.3)
DHEA-S (ng/ml)
197 (149–391)
230 (149–474)
Figures indicate mean ± SD or median (interquartile range) FSH follicle-stimulating hormone, LH luteinizing hormone, DHEA-S dehydroepiandrosterone sulfate Alive vs. dead * p \ 0.05 ** p \ 0.01
Discussion Our results showed that the prevalence of hypogonadism was present in about half (53.3 %) of male aged patients admitted for acute disease. This hypogonadism was mainly (56.3 %) non-hypergonadotropic, i.e., normo- or hypogonadotropic. The main cause of hospitalization in hypogonadic patients was respiratory tract infection, which was present in about one quarter (23.7 %) of hypogonadic
J Endocrinol Invest (2014) 37:135–141
139
Table 4 Results of three models of logistic regression analysis used to assess the influence of clinical parameters (age and BMI) and several hormones on the mortality during hospitalization in elderly hospitalized patients Covariates
Model 1
Model 2
Model 3
Odds ratio (95 % C.I.)
p
Odds ratio (95 % C.I.)
p
Odds ratio (95 % C.I.)
p
Testosterone (ng/dl)
0.99 (0.98–0.99)
0.008
0.99 (0.98–0.99)
0.019
0.99 (0.98–0.99)
0.021
Delta 4-androstenedione (ng/ml)
2.29 (1.07–4.91)
0.033
2.48 (1.13–5.44)
0.023
2.64 (1.00–7.02)
0.051
PRL (ng/ml) LH (mU/ml)
1.02 (0.99–1.06) 1.02 (0.97–1.07)
0.164 0.478
1.02 (0.99–1.06) 1.02 (0.97–1.07)
0.175 0.482
FSH (mU/ml)
0.98 (0.90–1.05)
0.545
0.98 (0.90–1.06)
0.569
Age (year)
1.03 (0.90–1.18)
0.661
1.03 (0.89–1.18)
0.700
BMI (kg/m2)
0.98 (0.84–1.15)
0.842
DHEA-S (ng/ml)
1.00 (1.00–1.00)
0.823
Estradiol (pg/ml)
1.00 (0.97–1.02)
0.853
Bold values indicate statistical significance PRL prolactin, LH luteinizing hormone, FSH follicle-stimulating hormone, BMI body mass index, DHEA-S dehydroepiandrosterone sulfate
Fig. 2 Probability of death during hospitalization in aged hospitalized male patients as a function of serum concentrations of total testosterone at the time of hospital admission, according to a model of logistic regression analysis
patients. Hypogonadism was also clearly associated to mortality during hospital stay and low serum concentrations of T was one of the main predictors of mortality during hospitalization in this population. Serum T concentrations decrease from fifth decade of life showing a high percentage of men over 60 years with T concentrations below the lower limit of normal range for a young adult male (20–30 years) [7]. Prevalence of hypogonadism in men varies with the definition used or cutpoints employed, the population studied, and method of analysis used [5]. In older men the prevalence of hypogonadal T levels increased to about 20 % of men over 60, 30 % over 70 and 50 % over 80 year of age, and even greater percentages
when free T index (free T index = T/SHBG) criteria were used [2, 3, 5]. In our study, the prevalence of hypogonadism was as expected (*50 %) according to the elevated age of the patients (mean age 86 years). Other factors as the presence of common acute and chronic illnesses can increase the prevalence of hypogonadism in men [22]. It is possible that the coexistence of acute and chronic illnesses at the time of the study in our population could have positively influenced on the elevated prevalence. Aging is accompanied by insidious depressant changes in hypothalamic–pituitary–gonadal axis function in men [4]. Among main postulated defects are: (1) reduced gonadotropin-releasing hormone (GnRH) pulse size, (2) increased GnRH pulse number, (3) decreased endogenous T negative feedback, (4) small amplitude of LH pulses, (5) high serum SHBG concentration, (6) impaired action of LH pulses on T synthesis and, (7) decreased androgen receptors [23]. That is why several patterns of gonadotropin secretion in the elderly can be found. Moreover, many aged individuals have often multiple chronic conditions and many of them are usually polymedicated. All these factors may adversely affect gonadal function at different levels. Our study showed that hypogonadism was predominantly (56.3 %) non-hypergonadotropic. Furthermore, 15 % of hypogonadic patients were hypogonadotropic versus none in the eugonadal group. These findings suggest that central (hypothalamic or pituitary) failure is more prevalent than primary gonadal hypofunction in the patophysiology of hypogonadism in hospitalized aged males. The relationship between serum T concentrations and vital status in the elderly male patient at admission was established by demonstrating a negative correlation with age and a positive correlation with parameters related to the level of independence of the patient as Barthel index [20].
123
140
An inverse relationship between serum T concentration and cardiovascular risk has been reported in the elderly [12, 13], but not always [16]. Lower T levels have been associated with insulin resistance, metabolic syndrome, diabetes, carotid intima-media thickness, and aortic and lower limb arterial disease in men, predisposing to increased cardiovascular risk [12, 13, 19]. In our study, we found not only an association between hypogonadism and personal history of heart disease but also with cancer, respiratory pathology, and renal insufficiency, suggesting a non-specific increased comorbidity risk associated with low serum T concentrations in this population. We found a relationship between lower T concentrations and respiratory tract infection. In fact, respiratory tract infection was 1.8 times more frequent in hypogonadal than in eugonadal patients. The reason for this association in not clear; however, severe hypoxemia and hypercapnia, more than cytokines associated to infection, could influence on the reduction level of serum T, as it has been previously reported in patients chronic obstructive pulmonary disease during stable and exacerbation phases [24]. Congestive heart failure was also more frequent (1.5 times) in hypogonadal compared to eugonadal patients. This might be in part a reflection of the association between hypogonadism and personal history of heart disease previously mentioned. Lastly, in contrast to other studies showing a higher incidence of stroke or transient ischemic attack (TIA) in communitydwelling older men with total testosterone \337 ng/dl (11.7 nmol/l) [25], we could not find this association, but rather the contrary, those patients who were admitted for acute cerebrovascular disease showed the highest serum T concentrations [466 ng/dl (195–610 ng/dl)] compared to other causes of hospital admission. The explanation of this finding is uncertain, although different factors such as advancing age of the patients, associated comorbidities, and higher prevalence of other causes of hospitalization, and the smaller number of patients and lower number of events seen in the study could explain these differences. Reduced T concentrations have been associated with higher cardiovascular and overall mortality in older men [12, 13, 17–19]. Furthermore, the increase in mortality has been related not only to cardiovascular disease, but also to cancer [26] and respiratory conditions [27]. In addition, in a recent observational study of mortality in men with low T levels, T treatment was associated with decreased mortality compared with no T treatment [28]. All these findings seem to prove the relationship between hypogonadism and mortality due to several causes in men. Our study confirms the association between low T levels and mortality during hospitalization in elderly hospitalized
123
J Endocrinol Invest (2014) 37:135–141
men due to acute illness (Fig. 1). Moreover, low T level was the most important predictive hormonal factor for hospital mortality in aged men, even after adjusting for confounders as age and BMI (Table 4). The relationship between low serum T levels and inhospital mortality in this population does not necessarily imply that hypogonadism per se could worsen the vital prognosis during hospitalization. It is possible that a low T level reflects a state of poor health, and therefore behaves as a marker of health as do other biological markers such as albumin, glucose, and triiodothyronine [29, 30]. Therefore, it does not seem appropriate to think that hormone replacement therapy with exogenous T is able to reverse a situation of vital risk in hospitalized aged patients. An incidental finding of our study was the positive predictor value of serum D-4 concentrations for in-hospital mortality. It is known that conversion of D-4 to testosterone requires the enzyme 17b-hydroxysteroid dehydrogenase (17b-HSD). It is possible that a decreased enzymatic activity of 17b-HSD associated to severe chronic pathological conditions and/or acute illness can influence the serum T and D-4 concentrations in aged hospitalized men. However, it must be emphasized that the statistical significance after adjusting for age and BMI was lost. It has been recently reported that hypogonadism-associated CV risk depends on the characteristics of subjects, being more evident in normal weight than in obese patients [31]. In our study, patients with low/normal weight (BMI \25 kg/m2) who died during hospitalization showed significant lower serum T concentrations than those who survived. This finding was not observed in overweight patients (BMI C25 kg/m2). However, we did not find a relationship between mortality due to major CV events and hypogonadic lean subjects indicating that serum T might be an unspecific marker of mortality in elderly population admitted by acute disease, although the number of patients who died was very small. In conclusion, our results show that alterations in gonadal function are present in about half of the elderly men admitted for acute illness, mainly due to non-hypergonadotropic hypogonadism. Respiratory tract infection is the main cause of admission in hypogonadic patients. Lastly, low T concentration is one of the best independent hormonal predictors of hospital mortality in this population. Funding Junta de Castilla y Leo´n (Spain), Consejerı´a de Sanidad, Proyectos de Investigacio´n en Biomedicina, Biotecnologı´a y Ciencias de la Salud. Expediente n° GRS 393/A/09. Conflict of interest The authors P. Iglesias, F. Prado, M.C. Macı´as, M.T. Guerrero, A. Mun˜oz, E. Ridruejo, P. Tajada, C. Garcı´a-Are´valo, and J.J. Dı´ez declare that they have no conflict of interest.
J Endocrinol Invest (2014) 37:135–141
141
References 1. Tostain JL, Blanc F (2008) Testosterone deficiency: a common, unrecognized syndrome. Nat Clin Pract Urol 5:388–396 2. Harman SM, Metter EJ, Tobin JD, Pearson J, Blackman MR, Baltimore Longitudinal Study of Aging (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 3. Vermeulen A, Kaufman JM (2002) Diagnosis of hypogonadism in the aging male. Aging Male 5:170–176 4. Allan CA, McLachlan RI (2004) Age-related changes in testosterone and the role of replacement therapy in older men. Clin Endocrinol (Oxf) 60:653–670 5. Araujo AB, O’Donnell AB, Brambilla DJ et al (2004) Prevalence and incidence of androgen deficiency in middle-aged and older men: estimates from the Massachusetts male aging study. J Clin Endocrinol Metab 89:5920–5926 6. Araujo AB, Esche GR, Kupelian V et al (2007) Prevalence of symptomatic androgen deficiency in men. J Clin Endocrinol Metab 92:4241–4247 7. Wang C, Nieschlag E, Swerdloff R et al (2008) Investigation, treatment and monitoring of late-onset hypogonadism in males: ISA, ISSAM, EAU, EAA and ASA recommendations. Eur J Endocrinol 159:507–514 8. Wu FC, Tajar A, Pye SR et al (2008) Hypothalamic-pituitarytesticular axis disruptions in older men are differentially linked to age and modifiable risk factors: the European male aging study. J Clin Endocrinol Metab 93:2737–2745 9. Venkat K, Desai M, Arora MM, Singh P, Khatkhatay MI (2009) Age-related changes in sex steroid levels influence bone mineral density in healthy Indian men. Osteoporos Int 20:955–962 10. Matsumoto AM (2002) Andropause: clinical implications of the decline in serum testosterone levels with aging in men. J Gerontol A Biol Sci Med Sci 57:M76–M99 11. Travison TG, Araujo AB, Kupelian V, O’Donnell AB, McKinlay JB (2007) The relative contributions of aging, health, and lifestyle factors to serum testosterone decline in men. J Clin Endocrinol Metab 92:549–555 12. Yeap BB (2010) Androgens and cardiovascular disease. Curr Opin Endocrinol Diabetes Obes 17:269–276 13. Tirabassi G, Gioia A, Giovannini L et al (2013) Testosterone and cardiovascular risk. Intern Emerg Med 8(Suppl 1):S65–S69 14. Araujo AB, Kupelian V, Page ST, Handelsman DJ, Bremner WJ, McKinlay JB (2007) Sex steroids and all-cause and cause-specific mortality in men. Arch Intern Med 167:1252–1260 15. Cummings-Vaughn LA, Malmstrom TK, Morley JE, Miller DK (2011) Testosterone is not associated with mortality in older African–American males. Aging Male 14:132–140 16. Haring R, Teng Z, Xanthakis V et al (2013) Association of sex steroids, gonadotrophins, and their trajectories with clinical
17.
18.
19.
20.
21.
22. 23.
24.
25.
26.
27.
28.
29.
30.
31.
cardiovascular disease and all-cause mortality in elderly men from the Framingham heart study. Clin Endocrinol (Oxf) 78:629–634 Lehtonen A, Huupponen R, Tuomilehto J et al (2008) Serum testosterone but not leptin predicts mortality in elderly men. Age Ageing 37:461–464 Tivesten A, Vandenput L, Labrie F et al (2009) Low serum testosterone and estradiol predict mortality in elderly men. J Clin Endocrinol Metab 94:2482–2488 Corona G, Rastrelli G, Monami M et al (2011) Hypogonadism as a risk factor for cardiovascular mortality in men: a meta-analytic study. Eur J Endocrinol 165:687–701 Shah S, Vanclay F, Cooper B (1989) Improving the sensitivity of the Barthel Index for stroke rehabilitation. J Clin Epidemiol 42:703–709 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 Kalyani RR, Gavini S, Dobs AS (2007) Male hypogonadism in systemic disease. Endocrinol Metab Clin N Am 36:333–348 Veldhuis JD, Keenan DM, Liu PY, Iranmanesh A, Takahashi PY, Nehra AX (2009) The aging male hypothalamic-pituitary-gonadal axis: pulsatility and feedback. Mol Cell Endocrinol 299:14–22 Karadag F, Ozcan H, Karul AB, Yilmaz M, Cildag O (2009) Sex hormone alterations and systemic inflammation in chronic obstructive pulmonary disease. Int J Clin Pract 63:275–281 Yeap BB, Hyde Z, Almeida OP et al (2009) Lower testosterone levels predict incident stroke and transient ischemic attack in older men. J Clin Endocrinol Metab 94:2353–2359 Khaw KT, Dowsett M, Folkerd E et al (2007) Endogenous testosterone and mortality due to all causes, cardiovascular disease, and cancer in men: European prospective investigation into cancer in Norfolk (EPIC-Norfolk) prospective population study. Circulation 116:2694–2701 Laughlin GA, Barrett-Connor E, Bergstrom J (2008) Low serum testosterone and mortality in older men. J Clin Endocrinol Metab 93:68–75 Shores MM, Smith NL, Forsberg CW, Anawalt BD, Matsumoto AM (2012) Testosterone treatment and mortality in men with low testosterone levels. J Clin Endocrinol Metab 97:2050–2058 Iglesias P, Mun˜oz A, Prado F et al (2009) Alterations in thyroid function tests in aged hospitalized patients: prevalence, aetiology and clinical outcome. Clin Endocrinol (Oxf) 70:961–967 Iglesias P, Polini A, Mun˜oz A et al (2011) Fasting hyperglycaemia and in-hospital mortality in elderly population. Int J Clin Pract 65:308–313 Corona G, Rastrelli G, Monami M et al (2011) Body mass index regulates hypogonadism-associated CV risk: results from a cohort of subjects with erectile dysfunction. J Sex Med 8:2098–2105
123
ORIGINAL ARTICLE
Hypogonadism in aged hospitalized male patients: prevalence and clinical outcome P. Iglesias • F. Prado • M. C. Macı´as • M. T. Guerrero • A. Mun˜oz • E. Ridruejo • P. Tajada • C. Garcı´a-Are´valo • J. J. Dı´ez
Received: 3 September 2013 / Accepted: 17 November 2013 / Published online: 9 January 2014 Ó Italian Society of Endocrinology (SIE) 2013
Abstract Objective Male hypogonadism is common in the elderly and has been associated with increased risk of mortality. Our objective has been to assess the prevalence of primary and central hypogonadism in elderly male patients admitted to the hospital because of acute illness. We also evaluated the relationships between gonadal dysfunction and in-hospital mortality. Patients and methods 150 patients, aged C65 years, admitted during 2010 and 2011 in our geriatric unit, were studied. Serum concentrations total, bioavailable and free testosterone, as well as of follicle-stimulating hormone and luteinizing hormone were quantified in every patient. Hypogonadism was defined by the presence of serum testosterone levels lower than 200 ng/dl. Results Hypogonadism was found in 80 patients (53.3 %). Serum gonadotropin concentrations were elevated in 43.7 % of these patients, whereas 41.3 % of hypogonadic patients showed normal and 15 % low gonadotropin concentrations. Respiratory tract infection and congestive heart failure were the main causes of hospitalization in hypogonadal men, whereas acute cerebrovascular disease was the main reason for admission in eugonadal patients. Of the 13 patients who died during hospitalization, 12 were hypogonadic. Patients
P. Iglesias (&) J. J. Dı´ez Department of Endocrinology, Hospital Ramo´n y Cajal, Ctra. de Colmenar, Km 9,100, 28034 Madrid, Spain e-mail: [email protected] F. Prado M. C. Macı´as M. T. Guerrero A. Mun˜oz E. Ridruejo Department of Geriatrics, Hospital General, Segovia, Spain P. Tajada C. Garcı´a-Are´valo Department of Biochemistry, Hospital General, Segovia, Spain
who died showed significantly lower serum levels of total, free and bioavailable testosterone than those found in patients who survived. Conclusion Our results show that about half of patients admitted for acute illness have hypogonadism, mainly of non-hypergonadotropic type. Gonadal hypofunction is significantly related with in-hospital mortality. A low value of serum testosterone may be a predictor for mortality in elderly male patients. Keywords Aged male patients Elderly Testosterone Hypogonadism Clinical outcome
Introduction Epidemiological surveys have shown that the prevalence of hypogonadism ranges from 6 to 9.5 % in communitydwelling men aged 40–70 years, and rises to 15–30 % in diabetic or obese men [1]. Due to the fact that gonadal function declines gradually with age in men, testosterone (T) deficiency is even more prevalent in the elderly ranging from 30 to 50 % [2–9]. An association between low T levels and morbidity and mortality in elderly men has been reported. Several chronic disease including diabetes mellitus, chronic obstructive pulmonary disease, inflammatory arthritis, kidney failure and HIV infection, obesity, metabolic syndrome and hemochromatosis are well-known risk factors for hypogonadism in the elderly [7, 10, 11]. On the other hand, a low serum T concentration has been considered as a cardiovascular risk factor in the elderly [12, 13]. Lastly, several large studies, although not all [14–16], have found a relationship between low T levels and cardiovascular and overall mortality in elderly men [12, 13, 17–19].
123
136
To the best of our knowledge, the prevalence and clinical significance of gonadal dysfunction in elderly male patients admitted for acute illness has not clearly been evaluated. We performed a cross-sectional study in a cohort of elderly male patients admitted to our hospital for acute disease. Our aim has been to assess the prevalence of hypogonadism in this population, as well as to study the relationship between gonadal derangements and morbidity and in-hospital mortality in these patients.
Patients and methods
J Endocrinol Invest (2014) 37:135–141
Diagnostics Inc., Tarrytown, NY, USA). The sensitivity (S) of the assay for these hormones was 0.3 mU/ml, 0.07 mU/ml, 0.3 ng/ml, 10 pg/ml, and 10 ng/dl, respectively. Normal ranges (NRs) for adult males were: FSH 1.4–18.1 mU/ml; LH 1.5–9.3 mU/ml; PRL 2.1–17.7 ng/ml; E2 11.6–41.2 pg/ml; and T 240–820 ng/dl. Other serum hormones analyzed by a chemiluminescent immunometric assay (Immulite 2,000 analyzer, Siemens Medical Solutions Diagnostics, Llanberis, Gwynedd, UK) were D-4 (NR 0.6–3.1 ng/ml; S 0.3 ng/ml), and DHEA-S (NR 80–560 mcg/dl; S 3 mcg/dl). Finally, SHBG was quantified by the same procedure and NR and S were 10–57 and 0.02 nmol/l, respectively.
Patients Criteria for diagnosis A cross-sectional study including all male patients older than 65 years that were hospitalized during two consecutive years (2010 and 2011) was carried out. All patients admitted to our geriatric unit were included in the study regardless of the cause of hospitalization. A detailed clinical history and complete physical examination were performed in every patient. Functional status at entry was evaluated using the modified version of Barthel index [20]. This index evaluates ten basic activities of everyday life with a score that ranges from 0 (maximal dependence) to 100 (maximal independence). Previous comorbidity and main cause for hospitalization were also registered and related to gonadal dysfunction. All deaths during hospitalization were registered. This study was approved by our local ethical committee and informed consent was given by subjects prior to participation in the study.
Patients were classified into eugonadal and hypogonadal (hyper-, normo- or hypogonadotropic) depending on the values of serum T and gonadotropins (FSH and LH). In this elderly population, patients were considered eugonadal when serum T concentrations were C200 ng/dl regardless of the gonadotropins values. The diagnosis of primary hypogonadism (hypergonadotropic) was established when patients simultaneously showed serum T concentrations \200 ng/dl and FSH and LH concentrations exceeding the upper limit of the normal reference range for these hormones ([18.1 mU/ml for FSH and [9.3 mU/ml for LH). Central or secondary hypogonadism (normo or hypogonadotropic) was considered when reduced serum T concentrations (\200 ng/dl) and normal or decreased gonadotropins levels (B18.1 mU/ml for FSH and B9.3 mU/ml for LH) were found.
Hormonal gonadal study
Statistical analysis
Gonadal function was assessed by measuring serum concentrations of gonadotropins (follicle-stimulating hormone, FSH and luteinizing hormone, LH) and total testosterone (T). Bioavailable (BioT) and free testosterone (FT) were calculated based on the values of T, sex hormone binding globulin (SHBG), and albumin, as previously described [21]. Other serum hormones analyzed were estradiol (E2), prolactin (PRL), delta-4-androstenedione (D-4), and dehydroepiandrosterone sulfate (DHEA-S).
For quantitative variables, results are expressed as mean ± SD for normally distributed data, and as median (range) for nonparametric data. Adjustment to normal distribution was tested by the Kolmogorov test. Categorical variables are described as percentages. For comparisons of means between two groups of subjects the Student t test was used for normally distributed data, and the Mann– Whitney test was employed for nonparametric data. For ratio comparisons the Chi square test and Fisher’s exact test were used. Correlations between quantitative variables were assessed using Spearman’s correlation analysis. Multivariate regression analysis between serum testosterone (dependent variable) and clinical and analytical variables was also evaluated. Models of logistic regression analysis were used to assess hospital mortality as a function of the different hormones analyzed. Differences were considered significant when p \ 0.05.
Hormone assays Fasting samples of venous blood were obtained from an antecubital vein between 08:30 and 09:00 h in the first day of admission for quantification. Serum FSH, LH, PRL, E2, and T were measured by a chemiluminiscence immunoassay (ADVIA Centaur XP analyzer, Siemens Healthcare
123
J Endocrinol Invest (2014) 37:135–141
137
Table 1 Clinical and analytical data of the 150 studied patients at hospital admission Mean ± SD or median (interquartile range)
Range (min/ max)
Age (years)
86.0 ± 4.8
68–99
Barthel index at admission
80 (50–90)
0–100
BMI (kg/m2)
25.7 ± 4.0
13–46
SBP (mmHg)
128 ± 24
60–193
DBP (mmHg)
70 ± 14
30–113
Clinical data
Analytical data FSH (mU/ml)
8.2 (4.9–14.0)
0.7–81.3
LH (mU/ml)
9.0 (5.1–14.7)
0.06–80.2
PRL (ng/ml)
9.5 (7.0–13.9)
1.1–96.9
Estradiol (pg/ml)
19.4 (6.9–32.0)
5.9–222.0
Testosterone (ng/dl)
190 (90–330)
10–970
SHBG (nmol/l)
68 (34–67)
13–146
Albumin (g/dl) Free testosterone (ng/dl)
3.4 ± 0.5 3.2 (1.4–5.3)
1.6–4.6 0.1–16.0
Free testosterone (%)
1.7 ± 0.7
0.6–3.5
Bioavailable testosterone (ng/ dl)
58.5 (28.4–105.7)
2.1–363
Table 2 Distribution of the patients according to serum concentrations of gonadotropins and gonadal status Hypogonadism (Serum testosterone \200 ng/dl) Yes
Total
No
FSH Low
3 (3.8)
0
3 (2.0)
Normal
60 (75.0)
58 (82.9)
118 (78.7)
High
17 (21.2)
12 (17.1)
29 (19.3)
Low
12 (15.0)
0
12 (8.0)
Normal
33 (41.2)
36 (51.4)
69 (46.0)
High
35 (43.8)
34 (48.6)
69 (46.0)
LH**
FSH and LH* Both low
3 (3.7)
0
3 (2.0)
Both normal
33 (41.3)
35 (50.0)
68 (45.3)
Both high
17 (21.3)
11 (15.7)
28 (18.7)
FSH normal/LH low
9 (11.3)
0
9 (6.0)
FSH normal/LH high LH normal/FSH low
18 (22.4) 0
23 (32.9) 0
41 (27.3) 0
0
1 (1.4)
1 (0.7)
80 (53.3)
70 (46.7)
150
LH normal/FSH high Total
Bioavailable testosterone (%)
31.9 ± 10.6
10.4–63.1
Figures indicate the number of patients (%)
Delta 4-androstenedione (ng/ ml)
1.0 (0.6–1.4)
0.02–4.6
Hypogonadism vs. normal testosterone
DHEA-S (mcg/dl)
220 (149–455)
28–2030
* p = 0.014
BMI body mass index, SBP systolic blood pressure, DBP diastolic blood pressure, LH luteinizing hormone, FSH follicle-stimulating hormone, PRL prolactin, SHBG sex hormone binding globulin, DHEA-S dehydroepiandrosterone sulfate
Results Clinical features and analytical data A group of 150 male patients (mean age ± SD, 86.0 ± 4.8 years) was studied. The main causes of admission at the hospital were: congestive heart failure (28 patients, 18.7 %), acute cerebrovascular disease (21 patients, 14.0 %), respiratory tract infection (28 patients, 18.7 %), acute digestive hemorrhage (13 patients, 8.7 %), coronary heart disease (8 patients, 5.3 %), exacerbation of chronic obstructive pulmonary disease (7 patients, 4.7 %), cancer (4 patients, 2.7 %), anemia (3 patients, 2.0 %), and other (38 patients, 25.2 %). Clinical and hormonal data of the studied patients are summarized in Table 1. Prevalence of gonadal dysfunction Hypogonadism was present in about half of the patients (80 patients, 53.3 %). Among them, 43.7 % were
** p = 0.003
hypergonadotropic, 41.3 % normogonadotropic and 15 % hypogonadotropic. Seventy (46.7 %) patients were eugonadal, half of them (35 patients, 50 %) normogonadotropic and the other half hypergonadotropic. We found a statistically significant (p = 0.003) difference in the proportion of hypo-, normo- and hypergonadotropic patients between patients with and without hypogonadism, mainly due to the presence of a greater percentage of hypogonadotropic patients in the hypogonadal group. Table 2 shows the distribution of the patients according to the level of serum gonadotropins and the presence or absence of hypogonadism. Regression analysis In the bivariate correlation analysis, serum T concentration was positively correlated with Barthel index (q = 0.280, p = 0.001), systolic blood pressure (SBP) (q = 0.212, p = 0.009), albumin (q = 0.334, p \ 0.001), SHBG (q = 0.265, p = 0.001), E2 (q = 0.405, p \ 0.001), and D-4 (q = 0.318, p \ 0.001). On the contrary, T was negatively correlated only with age (q = -0.301, p \ 0.001). Multivariate regression analysis showed a significant correlation between serum T concentration (dependent
123
138
J Endocrinol Invest (2014) 37:135–141
variable) and E2 (b = 0.187, p = 0.018), albumin (b = 0.234, p = 0.002), and SHBG (b = 0.270, p = 0.000). Gonadal function and associated diseases There was an association between hypogonadism and personal history of heart disease (heart failure, angina pectoris and/or acute myocardial infarction) (p \ 0.032), cancer (p = 0.004), respiratory pathology (p = 0.016), and renal insufficiency (p = 0.013). However, we could not find any association between hypogonadal status and hypertension, diabetes, hyperlipidemia, anemia, and prostate pathology. A significant relationship between gonadal status and the main cause of hospitalization was also found (p = 0.034). In this regard, the three main causes of admission for hypogonadic patients were respiratory tract infection (23.7 %), congestive heart failure (22.5 %), and acute cerebrovascular disease (6.2 %); however, in the eugonadal group these percentages were 12.8, 14.3 and 22.8 %, respectively. Moreover, patients admitted by respiratory tract infection and acute cerebrovascular disease showed the lowest and the highest values of T concentration [103 ng/dl (44–236 ng/dl) vs. 466 ng/dl (195–610 ng/dl), p \ 0.05]. Gonadal function and mortality Thirteen patients (8.7 %) died during hospitalization. Hypogonadism was associated to mortality during hospital stay. In fact, most of the patients (12 out of 13) who died were hypogonadic compared with 49.6 % of the patients who survived (p = 0.003) (Fig. 1). Patients who died showed significantly lower serum T (p \ 0.01), FT (p \ 0.01), and BioT (p \ 0.01) than survivors (Table 3). In-hospital mortality was associated with a significantly lower serum T concentrations both in very old patients (C85 year) (p = 0.013) and in those with a body mass index (BMI) \25 kg/m2 (p = 0.005) but not in patients \85 year and in those with a BMI C25 kg/m2. Three models of logistic regression analysis performed to study the influence of assessed hormones and clinical parameters (age and BMI) on fatal outcome during hospitalization showed that serum levels of T and D-4 were independent predictors of mortality in hospitalized elderly men population (Table 4). However, after adjusting for age and BMI, only T, but not D-4, showed a significant relationship with hospital mortality [OR 0.993 (0.987–0.999), P = 0.02]. Lastly, according to a model of logistic regression analysis, the likelihood of death during hospitalization in elderly men increased at lower serum T concentrations, ranging between 0.07 and 0.23 (Fig. 2).
123
Fig. 1 Percent distribution of the patients according to gonadal status at entry and hospital mortality Table 3 Distribution of the patients according to vital status at discharge Hospital mortality Yes (n = 13, 8.7 %)
No (n = 137, 91.3 %)
FSH (mU/ml)
6.1 (3.0–20.2)
8.3 (5.1–14.0)
LH (mU/ml)
6.2 (5.0–27.6)
9.1 (5.1–14.2)
PRL (ng/ml)
11.9 (9.4–25.0)
9.3 (6.9–13.9)
Estradiol (pg/ml)
22.5 (11.0–32.2)
19.0 (6.9–32.1)
Testosterone (ng/dl)
89 (39–158)**
201 (104–369)
Free testosterone (ng/dl)
1.4 (0.5–2.3)**
3.6 (1.7–5.6)
Free testosterone (%)
1.74 ± 0.56
1.68 ± 0.57
Bioavailable testosterone (ng/dl) Bioavailable testosterone (%)
28.8 (5.8–45.1)**
68.8 (30.1–110.0)
30.2 ± 11.8
32.1 ± 10.5
Delta 4-androstenedione (ng/ml)
1.4 (0.6–2.0)
1.0 (0.6–1.3)
DHEA-S (ng/ml)
197 (149–391)
230 (149–474)
Figures indicate mean ± SD or median (interquartile range) FSH follicle-stimulating hormone, LH luteinizing hormone, DHEA-S dehydroepiandrosterone sulfate Alive vs. dead * p \ 0.05 ** p \ 0.01
Discussion Our results showed that the prevalence of hypogonadism was present in about half (53.3 %) of male aged patients admitted for acute disease. This hypogonadism was mainly (56.3 %) non-hypergonadotropic, i.e., normo- or hypogonadotropic. The main cause of hospitalization in hypogonadic patients was respiratory tract infection, which was present in about one quarter (23.7 %) of hypogonadic
J Endocrinol Invest (2014) 37:135–141
139
Table 4 Results of three models of logistic regression analysis used to assess the influence of clinical parameters (age and BMI) and several hormones on the mortality during hospitalization in elderly hospitalized patients Covariates
Model 1
Model 2
Model 3
Odds ratio (95 % C.I.)
p
Odds ratio (95 % C.I.)
p
Odds ratio (95 % C.I.)
p
Testosterone (ng/dl)
0.99 (0.98–0.99)
0.008
0.99 (0.98–0.99)
0.019
0.99 (0.98–0.99)
0.021
Delta 4-androstenedione (ng/ml)
2.29 (1.07–4.91)
0.033
2.48 (1.13–5.44)
0.023
2.64 (1.00–7.02)
0.051
PRL (ng/ml) LH (mU/ml)
1.02 (0.99–1.06) 1.02 (0.97–1.07)
0.164 0.478
1.02 (0.99–1.06) 1.02 (0.97–1.07)
0.175 0.482
FSH (mU/ml)
0.98 (0.90–1.05)
0.545
0.98 (0.90–1.06)
0.569
Age (year)
1.03 (0.90–1.18)
0.661
1.03 (0.89–1.18)
0.700
BMI (kg/m2)
0.98 (0.84–1.15)
0.842
DHEA-S (ng/ml)
1.00 (1.00–1.00)
0.823
Estradiol (pg/ml)
1.00 (0.97–1.02)
0.853
Bold values indicate statistical significance PRL prolactin, LH luteinizing hormone, FSH follicle-stimulating hormone, BMI body mass index, DHEA-S dehydroepiandrosterone sulfate
Fig. 2 Probability of death during hospitalization in aged hospitalized male patients as a function of serum concentrations of total testosterone at the time of hospital admission, according to a model of logistic regression analysis
patients. Hypogonadism was also clearly associated to mortality during hospital stay and low serum concentrations of T was one of the main predictors of mortality during hospitalization in this population. Serum T concentrations decrease from fifth decade of life showing a high percentage of men over 60 years with T concentrations below the lower limit of normal range for a young adult male (20–30 years) [7]. Prevalence of hypogonadism in men varies with the definition used or cutpoints employed, the population studied, and method of analysis used [5]. In older men the prevalence of hypogonadal T levels increased to about 20 % of men over 60, 30 % over 70 and 50 % over 80 year of age, and even greater percentages
when free T index (free T index = T/SHBG) criteria were used [2, 3, 5]. In our study, the prevalence of hypogonadism was as expected (*50 %) according to the elevated age of the patients (mean age 86 years). Other factors as the presence of common acute and chronic illnesses can increase the prevalence of hypogonadism in men [22]. It is possible that the coexistence of acute and chronic illnesses at the time of the study in our population could have positively influenced on the elevated prevalence. Aging is accompanied by insidious depressant changes in hypothalamic–pituitary–gonadal axis function in men [4]. Among main postulated defects are: (1) reduced gonadotropin-releasing hormone (GnRH) pulse size, (2) increased GnRH pulse number, (3) decreased endogenous T negative feedback, (4) small amplitude of LH pulses, (5) high serum SHBG concentration, (6) impaired action of LH pulses on T synthesis and, (7) decreased androgen receptors [23]. That is why several patterns of gonadotropin secretion in the elderly can be found. Moreover, many aged individuals have often multiple chronic conditions and many of them are usually polymedicated. All these factors may adversely affect gonadal function at different levels. Our study showed that hypogonadism was predominantly (56.3 %) non-hypergonadotropic. Furthermore, 15 % of hypogonadic patients were hypogonadotropic versus none in the eugonadal group. These findings suggest that central (hypothalamic or pituitary) failure is more prevalent than primary gonadal hypofunction in the patophysiology of hypogonadism in hospitalized aged males. The relationship between serum T concentrations and vital status in the elderly male patient at admission was established by demonstrating a negative correlation with age and a positive correlation with parameters related to the level of independence of the patient as Barthel index [20].
123
140
An inverse relationship between serum T concentration and cardiovascular risk has been reported in the elderly [12, 13], but not always [16]. Lower T levels have been associated with insulin resistance, metabolic syndrome, diabetes, carotid intima-media thickness, and aortic and lower limb arterial disease in men, predisposing to increased cardiovascular risk [12, 13, 19]. In our study, we found not only an association between hypogonadism and personal history of heart disease but also with cancer, respiratory pathology, and renal insufficiency, suggesting a non-specific increased comorbidity risk associated with low serum T concentrations in this population. We found a relationship between lower T concentrations and respiratory tract infection. In fact, respiratory tract infection was 1.8 times more frequent in hypogonadal than in eugonadal patients. The reason for this association in not clear; however, severe hypoxemia and hypercapnia, more than cytokines associated to infection, could influence on the reduction level of serum T, as it has been previously reported in patients chronic obstructive pulmonary disease during stable and exacerbation phases [24]. Congestive heart failure was also more frequent (1.5 times) in hypogonadal compared to eugonadal patients. This might be in part a reflection of the association between hypogonadism and personal history of heart disease previously mentioned. Lastly, in contrast to other studies showing a higher incidence of stroke or transient ischemic attack (TIA) in communitydwelling older men with total testosterone \337 ng/dl (11.7 nmol/l) [25], we could not find this association, but rather the contrary, those patients who were admitted for acute cerebrovascular disease showed the highest serum T concentrations [466 ng/dl (195–610 ng/dl)] compared to other causes of hospital admission. The explanation of this finding is uncertain, although different factors such as advancing age of the patients, associated comorbidities, and higher prevalence of other causes of hospitalization, and the smaller number of patients and lower number of events seen in the study could explain these differences. Reduced T concentrations have been associated with higher cardiovascular and overall mortality in older men [12, 13, 17–19]. Furthermore, the increase in mortality has been related not only to cardiovascular disease, but also to cancer [26] and respiratory conditions [27]. In addition, in a recent observational study of mortality in men with low T levels, T treatment was associated with decreased mortality compared with no T treatment [28]. All these findings seem to prove the relationship between hypogonadism and mortality due to several causes in men. Our study confirms the association between low T levels and mortality during hospitalization in elderly hospitalized
123
J Endocrinol Invest (2014) 37:135–141
men due to acute illness (Fig. 1). Moreover, low T level was the most important predictive hormonal factor for hospital mortality in aged men, even after adjusting for confounders as age and BMI (Table 4). The relationship between low serum T levels and inhospital mortality in this population does not necessarily imply that hypogonadism per se could worsen the vital prognosis during hospitalization. It is possible that a low T level reflects a state of poor health, and therefore behaves as a marker of health as do other biological markers such as albumin, glucose, and triiodothyronine [29, 30]. Therefore, it does not seem appropriate to think that hormone replacement therapy with exogenous T is able to reverse a situation of vital risk in hospitalized aged patients. An incidental finding of our study was the positive predictor value of serum D-4 concentrations for in-hospital mortality. It is known that conversion of D-4 to testosterone requires the enzyme 17b-hydroxysteroid dehydrogenase (17b-HSD). It is possible that a decreased enzymatic activity of 17b-HSD associated to severe chronic pathological conditions and/or acute illness can influence the serum T and D-4 concentrations in aged hospitalized men. However, it must be emphasized that the statistical significance after adjusting for age and BMI was lost. It has been recently reported that hypogonadism-associated CV risk depends on the characteristics of subjects, being more evident in normal weight than in obese patients [31]. In our study, patients with low/normal weight (BMI \25 kg/m2) who died during hospitalization showed significant lower serum T concentrations than those who survived. This finding was not observed in overweight patients (BMI C25 kg/m2). However, we did not find a relationship between mortality due to major CV events and hypogonadic lean subjects indicating that serum T might be an unspecific marker of mortality in elderly population admitted by acute disease, although the number of patients who died was very small. In conclusion, our results show that alterations in gonadal function are present in about half of the elderly men admitted for acute illness, mainly due to non-hypergonadotropic hypogonadism. Respiratory tract infection is the main cause of admission in hypogonadic patients. Lastly, low T concentration is one of the best independent hormonal predictors of hospital mortality in this population. Funding Junta de Castilla y Leo´n (Spain), Consejerı´a de Sanidad, Proyectos de Investigacio´n en Biomedicina, Biotecnologı´a y Ciencias de la Salud. Expediente n° GRS 393/A/09. Conflict of interest The authors P. Iglesias, F. Prado, M.C. Macı´as, M.T. Guerrero, A. Mun˜oz, E. Ridruejo, P. Tajada, C. Garcı´a-Are´valo, and J.J. Dı´ez declare that they have no conflict of interest.
J Endocrinol Invest (2014) 37:135–141
141
References 1. Tostain JL, Blanc F (2008) Testosterone deficiency: a common, unrecognized syndrome. Nat Clin Pract Urol 5:388–396 2. Harman SM, Metter EJ, Tobin JD, Pearson J, Blackman MR, Baltimore Longitudinal Study of Aging (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 3. Vermeulen A, Kaufman JM (2002) Diagnosis of hypogonadism in the aging male. Aging Male 5:170–176 4. Allan CA, McLachlan RI (2004) Age-related changes in testosterone and the role of replacement therapy in older men. Clin Endocrinol (Oxf) 60:653–670 5. Araujo AB, O’Donnell AB, Brambilla DJ et al (2004) Prevalence and incidence of androgen deficiency in middle-aged and older men: estimates from the Massachusetts male aging study. J Clin Endocrinol Metab 89:5920–5926 6. Araujo AB, Esche GR, Kupelian V et al (2007) Prevalence of symptomatic androgen deficiency in men. J Clin Endocrinol Metab 92:4241–4247 7. Wang C, Nieschlag E, Swerdloff R et al (2008) Investigation, treatment and monitoring of late-onset hypogonadism in males: ISA, ISSAM, EAU, EAA and ASA recommendations. Eur J Endocrinol 159:507–514 8. Wu FC, Tajar A, Pye SR et al (2008) Hypothalamic-pituitarytesticular axis disruptions in older men are differentially linked to age and modifiable risk factors: the European male aging study. J Clin Endocrinol Metab 93:2737–2745 9. Venkat K, Desai M, Arora MM, Singh P, Khatkhatay MI (2009) Age-related changes in sex steroid levels influence bone mineral density in healthy Indian men. Osteoporos Int 20:955–962 10. Matsumoto AM (2002) Andropause: clinical implications of the decline in serum testosterone levels with aging in men. J Gerontol A Biol Sci Med Sci 57:M76–M99 11. Travison TG, Araujo AB, Kupelian V, O’Donnell AB, McKinlay JB (2007) The relative contributions of aging, health, and lifestyle factors to serum testosterone decline in men. J Clin Endocrinol Metab 92:549–555 12. Yeap BB (2010) Androgens and cardiovascular disease. Curr Opin Endocrinol Diabetes Obes 17:269–276 13. Tirabassi G, Gioia A, Giovannini L et al (2013) Testosterone and cardiovascular risk. Intern Emerg Med 8(Suppl 1):S65–S69 14. Araujo AB, Kupelian V, Page ST, Handelsman DJ, Bremner WJ, McKinlay JB (2007) Sex steroids and all-cause and cause-specific mortality in men. Arch Intern Med 167:1252–1260 15. Cummings-Vaughn LA, Malmstrom TK, Morley JE, Miller DK (2011) Testosterone is not associated with mortality in older African–American males. Aging Male 14:132–140 16. Haring R, Teng Z, Xanthakis V et al (2013) Association of sex steroids, gonadotrophins, and their trajectories with clinical
17.
18.
19.
20.
21.
22. 23.
24.
25.
26.
27.
28.
29.
30.
31.
cardiovascular disease and all-cause mortality in elderly men from the Framingham heart study. Clin Endocrinol (Oxf) 78:629–634 Lehtonen A, Huupponen R, Tuomilehto J et al (2008) Serum testosterone but not leptin predicts mortality in elderly men. Age Ageing 37:461–464 Tivesten A, Vandenput L, Labrie F et al (2009) Low serum testosterone and estradiol predict mortality in elderly men. J Clin Endocrinol Metab 94:2482–2488 Corona G, Rastrelli G, Monami M et al (2011) Hypogonadism as a risk factor for cardiovascular mortality in men: a meta-analytic study. Eur J Endocrinol 165:687–701 Shah S, Vanclay F, Cooper B (1989) Improving the sensitivity of the Barthel Index for stroke rehabilitation. J Clin Epidemiol 42:703–709 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 Kalyani RR, Gavini S, Dobs AS (2007) Male hypogonadism in systemic disease. Endocrinol Metab Clin N Am 36:333–348 Veldhuis JD, Keenan DM, Liu PY, Iranmanesh A, Takahashi PY, Nehra AX (2009) The aging male hypothalamic-pituitary-gonadal axis: pulsatility and feedback. Mol Cell Endocrinol 299:14–22 Karadag F, Ozcan H, Karul AB, Yilmaz M, Cildag O (2009) Sex hormone alterations and systemic inflammation in chronic obstructive pulmonary disease. Int J Clin Pract 63:275–281 Yeap BB, Hyde Z, Almeida OP et al (2009) Lower testosterone levels predict incident stroke and transient ischemic attack in older men. J Clin Endocrinol Metab 94:2353–2359 Khaw KT, Dowsett M, Folkerd E et al (2007) Endogenous testosterone and mortality due to all causes, cardiovascular disease, and cancer in men: European prospective investigation into cancer in Norfolk (EPIC-Norfolk) prospective population study. Circulation 116:2694–2701 Laughlin GA, Barrett-Connor E, Bergstrom J (2008) Low serum testosterone and mortality in older men. J Clin Endocrinol Metab 93:68–75 Shores MM, Smith NL, Forsberg CW, Anawalt BD, Matsumoto AM (2012) Testosterone treatment and mortality in men with low testosterone levels. J Clin Endocrinol Metab 97:2050–2058 Iglesias P, Mun˜oz A, Prado F et al (2009) Alterations in thyroid function tests in aged hospitalized patients: prevalence, aetiology and clinical outcome. Clin Endocrinol (Oxf) 70:961–967 Iglesias P, Polini A, Mun˜oz A et al (2011) Fasting hyperglycaemia and in-hospital mortality in elderly population. Int J Clin Pract 65:308–313 Corona G, Rastrelli G, Monami M et al (2011) Body mass index regulates hypogonadism-associated CV risk: results from a cohort of subjects with erectile dysfunction. J Sex Med 8:2098–2105
123
