Serum testosterone levels using the radioimmunoassay method in healthy Japanese male volunteers

Abstract

Aim:  The objective of the present study was to measure serum free and total testosterone values using the radioimmunoassay (RIA) method in healthy Japanese male volunteers with no current diseases.

Methods:  Two hundred and fifty‐one healthy men who had no medical illness and received no current medical treatment were selected from 405 male volunteers. Free and total testosterone were measured in blood samples using the RIA method.

Results:  Free but not total testosterone significantly decreased with age. Mean free testosterone values from morning blood samples for each age decade from the 20s to the 70s were 17.0, 14.6, 12.5, 10.6, 8.9 and 8.5 pg/mL, respectively. Mean total testosterone values from morning blood samples for each age decade from the 20s to the 70s were 4.7, 4.2, 4.4, 4.2, 4.2, 4.0 and 4.0 ng/mL, respectively. The rates of healthy volunteers that fell within the standard reference ranges for free and total testosterone were 97% and 97%, respectively. However, 19% of the total testosterone values were considered to indicate hypogonadism according to the International Society for the Study of the Aging Male (ISSAM) criteria (<3.17 ng/mL).

Conclusions:  Our data corresponded to the standard reference ranges of Japanese men but not the ISSAM criteria. It may be more appropriate to establish a standard reference range for serum testosterone for individual countries. (Reprod Med Biol 2006; 5: 37–41)

INTRODUCTION

VARIOUS CLINICAL SYMPTOMS associated with age‐related androgen decline in men are known as partial androgen deficiency in the aging male (PADAM) or late onset hypogonadism. ¹ , ² , ³ Management of this syndrome will be one of the tasks for the medical field in our current aging society. For this purpose, proper evaluation of a patient's androgenicity is required. ⁴ The first step in this difficult task is to measure serum testosterone (T) level.

It is widely accepted that measurement of free and unbound T levels are accurate parameters of serum bioactive T level. ¹ , ⁵ However, these ideal parameters (i.e. free T measured using equilibrium dialysis, bioavailable T and calculated free T) are not always accessible in a clinical setting because of their cost and the time period needed for measurement. In Japan, these adverse situations allow access to limited T parameters, free and total T using radioimmunoassay (RIA) methods.

A recent report clearly demonstrated the clinical usefulness of free and total T measured using RIA; ⁶ however, until recently there were no reliable reference ranges for these T parameters in Japan. Iwamoto et al. ⁷ reported standard reference ranges for total and free T and these were recently adopted by the Japanese Urological Association (JUA). The present study provides reliable values based on a large number of subjects using well‐controlled methods, although the subjects came from different backgrounds.

It is well known that health status, mental condition and drug administration may affect serum T levels. ⁸ , ⁹ , ¹⁰ Thus, T values from healthy Japanese men, who have no current diseases physically or psychologically and no drug medication, may provide additional useful information.

Our goal was to measure free and total T using RIA methods in healthy male Japanese volunteers with no current diseases or drug medication. We compared our current data and the standard reference ranges reported by Iwamoto et al. and recommended by the JUA ⁷ with the criteria for hypogonadism of the International Society for the Study of the Aging Male (ISSAM) to examine their accordance with domestic and international criteria. ¹

MATERIALS AND METHODS

Volunteers

FIVE HOSPITALS IN Japan, Sanjukai Hospital (Sapporo), Zensyuukai Hospital (Maebashi), Kouchi‐Takasu Hospital (Kouchi), Fujisaki Hospital (Karatsu) and Niimura Hospital (Kagoshima), were involved in the present study. These five hospitals were located in different areas of Japan. Our hospital group recruited volunteers for this study. Four hundred and five male volunteers participated. We informed the volunteers of the objectives of the study before they agreed to participate. All volunteers gave blood samples and filled out a self‐administrated questionnaire. Blood samples were taken in the morning (8.00–11.00 hours) or afternoon (13.00–15.00 hours).

Definition of ‘healthy male’

We selected healthy male volunteers according to the following criteria based on answers to the self‐reported questionnaire. The definition of a ‘healthy male’ was:

• 1No current illness or medical treatment. Volunteers who did not have any current disease or a past history that might affect their T level and who were not receiving medical treatment were selected. The volunteers were asked questions about urological disease, liver disease, hormonal disease, hypertension, cardiovascular disease, respiratory disease, diabetes mellitus, cerebral infarction or bleeding, psychological disease and any cancer.
• 2No obesity because obesity can affect the serum T level. Volunteers with a body mass index (BMI) that was less than 30 were selected.
• 3Sense of well‐being. Volunteers with a sense of well‐being were selected.
• 4No need of support for activities of daily living (ADL). Volunteers who could carry out ADL without any help were selected.
• 5No depressive state. Volunteers who were not considered to be in a depressive state were selected according to two items of the first step screening of the Mini‐International Neuropsychiatric Interview (MINI) questionnaire. ¹¹ When the questionnaire was incomplete, we interviewed the volunteers directly.

Hormonal examination

All blood samples were measured for serum total and free T in the same laboratory, Mitsubishi BCL, Tokyo, using Diagnostic Products Corporation total and free testosterone kits (Diagnostic Products Corporation, Los Angeles, CA, USA). Hormonal values of healthy male volunteers were calculated by age decade to elicit any age‐associated changes in these hormones. Morning and afternoon blood samples were analyzed separately. The young adult mean (YAM) values of theses hormones for the 20s and 30s age decades were calculated based on hormonal values from the morning samples. All samples from this study were examined at five hospitals simultaneously from April 2003 to June 2003.

Low testosterone levels

The rates of volunteers who had a low serum T level were evaluated in the morning blood samples according to the standard reference ranges (mean ± 2 SD) of Japanese men reported by Iwamoto et al. ⁷ and the criteria for hypogonadism of ISSAM (3.17 ng/mL). ¹

Data analysis

The number of male volunteers from each hospital was 191 from Sanjyukai Hospital, 64 from Zensyukai Hospital, 50 from Kouchi‐Takasu Hospital, 50 from Fujisaki Hospital and 50 from Niimura Hospital. Two hundred and fifty‐one healthy male volunteers were selected from the initial volunteers according to the current criteria: 116 men from Sanjyukai Hospital, 42 from Zensyukai Hospital, 30 from Kouchi‐Takasu Hospital, 35 from Fujisaki Hospital and 28 from Niimura Hospital. Volunteers were selected at similar rates from all hospitals. Thus, we decided to combine the hormonal values of these healthy volunteers for the purposes of analysis.

One‐way anovas were used and P < 0.05 was considered significant.

RESULTS

Age‐associated changes in free T values from morning and afternoon blood samples

FREE T VALUES from morning and afternoon blood samples decreased significantly with age (Table 1). Free T values from morning and afternoon samples of subjects in their 20s were significantly higher than corresponding values from morning samples of men in their 40s, 50s, 60s and 70s and from afternoon samples of men in their 50s, 60s and 70s (Table 1). The only significant difference in free T values between morning and afternoon samples was found for men in their 20s and 30s (Table 1). The YAM value was 15.8 pg/mL.

Table 1.

Free testosterone (T) values in healthy Japanese male volunteers

Age (years)	Free T values for the morning sample (pg/mL) (mean ± SD) (n)	Free T values for the afternoon sample (pg/mL) (mean ± SD) (n)
20–29	17.0 ± 3.5 (16)†	13.0 ± 2.6 (12)
30–39	14.6 ± 3.1 (20)†	11.8 ± 3.6 (15)
40–49	12.5 ± 3.3 (20)‡	10.8 ± 2.3 (15)
50–59	10.6 ± 3.2 (39)‡	9.5 ± 2.3 (27)‡
60–69	8.9 ± 2.9 (34)‡	8.2 ± 2.8 (19)‡
70–79	8.5 ± 2.6 (20)‡	7.5 ± 3.0 (14)‡

^† Significantly higher free T value than the values for the afternoon samples in each age group.

^‡ ‡Significantly lower free T value than individuals in their 20s.

Age‐associated changes in total T values from morning and afternoon blood samples

Total T values from morning and afternoon samples did not show significant age‐related changes. The total T value of morning samples from subjects in their 20s and 30s were significantly higher than values from the corresponding afternoon samples (Table 2). The mean values of total cases and YAM were 4.19 and 4.58 ng/mL, respectively.

Table 2.

Total testosterone (T) values in healthy male volunteers

Age (years)	Total T values for the morning sample (pg/mL) (mean ± SD) (n)	Total T values for the afternoon sample (pg/mL) (mean ± SD) (n)
20–29	4.7 ± 1.4 (16)†	3.5 ± 1.2 (12)
30–39	4.2 ± 1.2 (20)†	3.6 ± 1.2 (15)
40–49	4.4 ± 1.3 (20)	4.0 ± 1.1 (15)
50–59	4.2 ± 1.4 (39)	3.7 ± 1.2 (27)
60–69	4.0 ± 1.4 (34)	3.4 ± 1.1 (19)
70–79	4.0 ± 1.4 (20)	3.5 ± 1.6 (14)
Total	4.2 ± 1.4 (149)	–

^† Significantly higher total T value than the values for the afternoon samples in each age group.

Comparison with Japanese standard reference ranges and international criteria of hypogonadism

The rates of free and total T values from morning samples that fell within the standard reference range reported by Iwamoto et al. ⁷ were 97% for both parameters (Table 3). In contrast to the good correspondence with the Japanese reference ranges, only 19% of the total T values were considered to indicate hypogonadism according to the ISSAM criteria (Table 4).

Table 3.

Rates of healthy volunteers within the standard reference ranges of free and total testosterone (T) reported by Iwamoto et al. ⁷

Age (years) (cases)	Rates (%) within the standard reference ranges of free T in each age decade (cases)	Rates (%) within the standard reference ranges of total T in each age decade (cases)
20–29 (16)	100 (16)	100 (16)
30–39 (20)	95 (19)	100 (20)
40–49 (20)	95 (19)	100 (20)
50–59 (39)	97 (38)	95 (37)
60–69 (34)	95 (32)	95 (32)
70–79 (20)	100 (20)	95 (19)
Total (149)	97 (144)	97 (144)

Table 4.

Rates of healthy volunteers considered to indicate hypogonadism according to the ISSAM criteria for total testosterone values

Age (years) (cases)	Rates (%) of hypogonadism (<3.17 ng/ml) (cases)
20–29 (16)	0 (0)
30–39 (20)	10 (2)
40–49 (20)	15 (3)
50–59 (39)	21 (8)
60–69 (34)	26 (9)
70–79 (20)	30 (6)
Total (149)	19 (28)

DISCUSSION

FREE T USING equilibrium dialysis but not RIA, bioavailable T (BT) and calculated free T (cFT) is recommended as an ideal parameter for serum T level. ¹ , ⁴ , ⁵ Measurement of free T using RIA is available, but its value tends to be lower than that measured using equilibrium dialysis. ⁵ However, recent studies have demonstrated that these ideal and practical parameters are well correlated. Tsujimura et al. clearly demonstrated that total and free T measured using RIA were significantly correlated with BT. ⁶ This result suggests that ‘clinical available T’ may reflect ‘bioavailable T’. Thus, total and free T measured using RIA are feasible and useful tools to screen and manage early and late onset of hypogonadism in countries where ideal parameters are not easily accessible.

Our first important finding was that our hormonal data corresponded well to the standard reference ranges of free and total T reported by Iwamoto et al. ⁷ This reference range provided long‐awaited data for many doctors who are involved in managing PADAM. Iwamoto et al.'s study was well organized and used sophisticated methods for the collection of blood samples and hormonal assays, which resulted in reliable data. However, the health condition of the subjects was unclear. Our current data, using only Japanese men who were considered healthy, support the reference ranges.

Reasons for the age‐related decline of T are primarily changes in testicular and hypothalamus–pituitary–axis functions. In addition, an increase in the prevalence of chronic illnesses may accelerate the age‐related decline of T. ⁸ , ⁹ A significant relationship between depressive symptoms and serum T level was reported in a community‐based study. ¹⁰ The current volunteers were considered to be healthy both physically and mentally, although health conditions were confirmed only with a self‐reported questionnaire, and the number of volunteers was not large. Thus, our data provide confirmative findings for the standard reference ranges of serum free and total T reported by Iwamoto et al. and additional useful and reliable information for the diagnosis and management of hypogonadism. Interestingly our data also correlated well with the free T values of Canadian data reported by Ooi et al. ¹⁴

In contrast to the good correlation between our current data and Iwamoto's and Ooi's data, only 19% of the volunteers were considered to suffer from hypogonadism according to the ISSAM criteria for total T values. A previous Korean study indicated a higher total T level than that recorded in our study. ¹² These differences may originate from race or methodology (i.e. the kits used for T measurements). These differences suggest that standard reference ranges of serum free and total T should be established in each country and include adjustments for the clinical situation of that country.

The diurnal rhythm of T secretion is imprecise in aged men. ¹³ Our data support this hypothesis. There was only a significant difference between the T levels in morning and afternoon blood samples in healthy volunteers in their 20s and 30s, although in general the T levels of afternoon blood samples tended to be lower than those of morning blood samples. The T values from afternoon blood samples may be useful for estimating a patient's T level, in particular for elderly men when morning blood sampling is not available. Although T values in morning samples are more reliable for diagnosis of hypogomadism, normal T values from afternoon blood samples may be helpful in excluding hypogonadism.

CONCLUSIONS

OUR DATA CORRESPONDED to the standard reference ranges of Japanese men but not to the ISSAM criteria. Thus, standard reference ranges of serum free and total T should be established for each country and include an adjustment for the clinical situation of the country.