Please Scroll Down to See Forums Below 
lanky
Well-known member
Clinically apparent prostate cancer occurs more commonly among Caucasians living in Western countries than in Chinese in the Far East. Prior studies demonstrated diminished facial and body hair and lower levels of plasma 3-androstanediol glucuronide and androsterone glucuronide in Chinese than in Caucasian men. Based upon these findings, investigators postulated that Chinese men could have diminished 5alpha-reductase activity with a resultant decrease in prostate tissue dihydrotestosterone levels and clinically apparent prostate cancer. An alternative hypothesis suggests that decreased 3-androstanediol glucuronide and androsterone glucuronide levels might reflect reduced production of androgenic ketosteroid precursors as a result of genetic or environmental factors. The present study examined 5-reductase activity, androgenic ketosteroid precursors, and the influence of genetic and environmental/dietary factors in groups of Chinese and Caucasian men. We found no significant differences in the ratios of 5ß-:5-reduced urinary steroids (a marker of 5-reductase activity) between Chinese subjects living in Beijing, China, and Caucasians living in Pennsylvania. To enhance the sensitivity of detection, we used an isotopic kinetic method to directly measure 5-reductase activity and found no difference in testosterone to dihydrotestosterone conversion ratios between groups. Then, addressing the alternative hypothesis, we found that the Caucasian subjects excreted significantly higher levels of individual and total androgenic ketosteroids than did their Chinese counterparts. To distinguish genetic from environmental/dietary factors as a cause of these differences, we compared Chinese men living in Pennsylvania and a similar group living in Beijing, China. We detected a reduction in testosterone production rates and total plasma testosterone and sex hormone-binding levels, but not in testosterone MCRs in Beijing Chinese as a opposed to those living in Pennsylvania. Comparing Pennsylvania Chinese with their Caucasian counterparts, we detected no significant differences in total testosterone, free and weakly bound testosterone, sex hormone-binding globulin levels, and testosterone production rates. Taken together, these studies suggest that environmental/dietary, but not genetic, factors influence androgen production and explain the differences between Caucasian and Chinese men.
Introduction
Top
Abstract
Introduction
Subjects and Methods
Results
Discussion
References
CHINESE men living in the Far East develop clinically apparent prostate cancer at a rate nearly 1/10th that in Caucasian men living in Western countries (1, 2, 3). The incidence of latent prostate cancer, on the other hand, is similar in the two populations. One potential explanation for this reduction in prostatic cancer risk in Chinese men is a genetically controlled diminution in 5-reductase activity. This would result in a lowering of prostate tissue dihydrotestosterone (DHT) levels and potentially of the rate of progression from latent to clinically apparent prostate cancer (4, 5). Alternatively, environmental or dietary factors that are prevalent in China but not in western countries could provide an explanation for the differences in cancer risk (6). These possibilities have led us to a series of studies comparing Chinese with Caucasian subjects.
We initially demonstrated that Chinese men have less facial and chest hair than age-matched Caucasian men and suggested the possibility that reduced conversion of testosterone to DHT in the Chinese might explain this difference (7, 8). In addressing this issue, we detected lower levels of the 5-reduced androgen metabolites, 3-androstanediol glucuronide (3-Diol-G) and 3-androsterone glucuronide (3A-G) in plasma of normal Chinese males and females compared to their Caucasian counterparts. It was apparent to us, however, that plasma levels of these conjugated metabolites provide only an indirect measure of tissue 5-reductase activity. The amount of circulating 3-Diol-G and 3-A-G is determined not only by tissue 5-reductase activity, but also by the amount of precursor ketosteroid secreted by the gonads and adrenals (9, 10). Accordingly, the present study used more direct measurements of 5-reductase activity and quantitative assessment of ketosteroids to evaluate differences between Chinese and Caucasian subjects. We chose techniques previously used to demonstrate 5-reductase deficiency in subjects with an inherited decrease in the activity of this enzyme. These methods involved measurement of the 5ß:5 ratios of several urinary metabolites using gas chromatography and mass spectrometry as well as isotopic kinetic determinations of the conversion of testosterone to DHT (9, 11).
The data reported here suggest that a genetic alteration of 5-reductase activity does not explain the reduced 3-Diol-G and 3-A-G levels in Chinese men. Instead, reduced levels of the androgenic ketosteroid precursors of these plasma metabolites provide the most likely explanation. Dietary or other environmental factors appear to alter the serum levels and production rates of testosterone of Chinese men living in China. These factors require consideration when postulating a causal link between the reduction of prostate cancer incidence, reduction of ketosteroid excretion, and lowered testosterone production rates in Chinese men living in China.
Subjects and Methods
Top
Abstract
Introduction
Subjects and Methods
Results
Discussion
References
Comparison of urinary metabolites in Caucasian and Chinese men and women
Overnight urine samples were obtained from 20 Caucasian men and 20 women living in the United States and from 20 Chinese men and 20 women living in Hong Kong. All subjects were normal medical students taking no medications that would alter androgen levels. Thirty-eight steroids were measured by gas chromatography-mass spectrometry (12). Assessment of 5-reductase activity, as previously described (9, 10), involved calculation of the ratios of four 5ß-:5-steroid pairs measured in overnight urine samples. For assessment of ketosteroid excretion, values were normalized by dividing by the urinary creatinine level. Total ketosteroid levels were calculated by adding all individual ketosteroid values together. These are expressed per mg creatinine to correct for variances in duration of collection of samples.
Measurement of androgens and sex hormone-binding globulin (SHBG) in Caucasian and Chinese men
Serum samples were obtained from an additional 10 Caucasian men (aged 22–27 yr) and 10 Chinese men (aged 20–37 yr) living in Pennsylvania and from 10 Chinese men (aged 24–39 yr) living in Beijing, China. Subjects living in Pennsylvania were examined at Pennsylvania State University Hospital and had a normal physical examination and no evidence of underlying illness. Subjects living in Beijing were examined by one investigator (G.-y.Z.) who noted that 9 of 10 had fathered children, and 1 had normal sexual function but was using condoms for contraception. No evidence of endocrine disorders or other illness was detected. Testis size was normal in all men, ranging from 15 mL (right and left testes) to 25 mL (right and left testes). Samples obtained from these subjects were used for measurement of testosterone, bioactive testosterone, and SHBG levels. All volunteers signed informed consents, and all protocols were approved as required by the institutions where the research was conducted. Total testosterone was measured by RIA as previously described (7, 8). Bioactive testosterone (free and weakly bound) was measured using the ammonium sulfate precipitation method (13, 14). SHBG was measured using a solid phase 125I immunoradiometric assay kit from Diagnostic Products Corp. (Los Angeles, CA).
Measurement of the interconversion of androgens in Caucasian and Chinese men
Eighty microcuries of [3H]testosterone were infused iv over a 3-h period after a 5-mL loading dose bolus of 2.5 µCi was given to the 30 male subjects. Identical infusion pumps and aliquots from centrally prepared radiolabeled testosterone were used for studies in the United States and China. Sixty milliliters of blood were collected after 2, 2.5, and 3 h of infusion. Serum was obtained and stored at -20 C until processed. Samples from Chinese men were flown to Pennsylvania frozen and processed identically to those collected in Pennsylvania. Upon thawing, the volume was measured, and [14C]testosterone, DHT, and androstenedione were added to each sample to determine recoveries. This was followed by 3 extractions with 200 mL methylene chloride. The methylene chloride fractions were combined, partially evaporated, backwashed twice with 3 mL distilled water, and taken to dryness. The samples were then purified by thin layer chromatography and high pressure liquid chromatography using solvent systems previously developed (15, 16) and outlined in Fig. 1. Calculations of conversion ratios and MCRs followed the methods of Mahoudeau et al. as extensively described previously (17).
View larger version (11K):
[in this window]
[in a new window]
Figure 1. Purification system to separate androgens used in this study. TLC, Thin layer chromatography; HPLC, high pressure liquid chromatography.
Data analysis
All samples were collected in the morning to minimize differences due to circadian variation in steroids. Due to the small number of individuals in each group and the known lack of normal distribution of hormonal data in these populations (18), data were expressed as median values with 25th and 75th percentiles in parentheses (in the tables) or were presented as box plots of the 25th and 75th percentiles with the median indicated by the line within the box and the range depicted by the bars outside the box (in the figures). Testing of statistical differences among groups used Kruksal-Wallis one-way ANOVA (when three groups were compared) and Mann-Whitney tests (when two groups were compared). All analyses were performed using the Systat computer package (Systat, Evanston, IL). The Bonferonni correction was used to compensate for multiple urine comparisons, and only P < 001 was considered statistically significant.
Results
Top
Abstract
Introduction
Subjects and Methods
Results
Discussion
References
Initial studies addressed the hypothesis that Chinese subjects have lower levels of 5-reductase activity than their Caucasian counterparts. The ratios of 5ß-:5-reduced urinary steroids were determined as a method to detect these differences. No significant differences in 5ß:5 ratios (or even suggestive trends) were found to suggest a diminution of 5-reductase activity in the Hong Kong Chinese compared to the Pennsylvania Caucasian medical students. (Table 1). Levels in the Chinese subjects were 6- to 30-fold lower than those in patients with known 5-reductase deficiency (9, 10) and were essentially the same as those in Caucasians. We next used the more sensitive isotopic kinetic methodology to enhance the likelihood of detecting small differences in 5-reductase activity (10, 11). To evaluate the possibility that dietary/environmental factors might confound interpretation, we compared Chinese living in Pennsylvania with Chinese living in Beijing and with Caucasians living in Pennsylvania. The percent conversion of testosterone to DHT as a direct measure of 5-reductase activity did not differ among the three groups studied (Fig. 2). As an additional marker of androgen metabolism, the conversion of testosterone to androstenedione also did not differ among these three groups of subjects (Fig. 2). Considering both the urinary and plasma isotopic data, we concluded that the Chinese subjects, whether living in China or in the United States, did not have lower 5-reductase activity than their Caucasian counterparts.
View this table:
[in this window]
[in a new window]
Table 1. Comparison of 5ß:5 ratios of urinary metabolites
View larger version (16K):
[in this window]
[in a new window]
Figure 2. Conversion of testosterone (T) to DHT or androstenedione (4A). Cau(U), Caucasians living in the U.S. Ch(U), Chinese living in the U.S. Ch(C), Chinese living in China. + indicates values lying greater than 1.5 times the difference between the 25th and 75th percentile values above or below the box. All statistical comparisons were nonsignificant. In the text, Beijing Chinese is frequently used in place of Ch(C), and Pennsylvania Chinese is used for Ch(U).
An alternative hypothesis suggested that the lower plasma 5-reduced metabolites in Chinese subjects than in their Caucasian counterparts might reflect a reduction of androgenic ketosteroid precursors as first proposed by Rittmaster and colleagues (19). To test this possibility, we measured a series of urinary ketosteroids, which serve as markers for adrenal androgen production. We detected lower levels of several of these components in the urines of Chinese compared to Caucasian subjects (Table 2). Androsterone and etiocholanolone were significantly lower in the Chinese subjects (data from males and females combined, P < 0.001 and P < 0.001, respectively), and 11ß-hydroxyetiocholanolone was marginally lower (P = 0.025). To integrate data from multiple steroid measurements, we also calculated the total level of excreted ketosteroids in each subject. With this methodology, we found total ketosteroid excretion to be significantly lower in Chinese than in Caucasian subjects (P = 0.006; Fig. 3). When examined separately by gender, the differences appeared to be greater in females than in males, but variances were substantial. As a control for potential confounding factors, such as the amount of creatinine excreted, we also measured glucocorticoid metabolites. These levels, in contrast, were not lower in Chinese than in Caucasian subjects (Table 2). Interestingly, three minor compounds, 6ß-hydroxycortisol, 20-dihydrocortisol, and 20ß-dihydrocortisol, were higher in the Chinese group (data from males and females combined, P = 0.044, 0.02, and 0.007, respectively), but not sufficiently to reach the Bonferonni limits of significance (i.e. P < 0.001).
View this table:
[in this window]
[in a new window]
Table 2. Urinary levels of individual ketosteroid and glucocorticoid metabolites
View larger version (18K):
[in this window]
[in a new window]
Figure 3. Ratios of total urinary ketosteroids (micrograms per dL) to creatinine (milligrams per dL) in Caucasian (Cau) and Chinese (CHI) subjects. + indicates outliers, as described in Fig. 2. Chinese subjects had significantly lower ketosteroids (*, P = 0.006). Individual comparisons between Caucasian and Chinese males and Caucasian and Chines females were not statistically significant.
As a potential explanation for the lowered ketosteroid excretion in subjects living in China, we postulated that environmental, dietary, or genetic effects could either act independently or in concert to alter androgen production or metabolism. Lacking ketosteroid data in Chinese men living in Pennsylvania (Pennsylvania Chinese), we instead analyzed isotopic kinetic data to compare androgen production and metabolism directly in Pennsylvania Chinese with Beijing, Chinese. Notably, we detected a reduction in testosterone production rates (Fig. 4) in the Beijing compared with the Pennsylvania Chinese (P = 0.034). In contrast, testosterone MCRs did not differ between these two groups (Fig. 4). We also detected significantly lower levels of total testosterone and SHBG levels (P = 0.031 and 0.038, respectively) in Beijing than in Pennsylvania Chinese (Fig. 5). The lower SHBG levels probably explain the inability to demonstrate lower bioavailable testosterone levels in the Chinese men living in Beijing than in those in Pennsylvania. Taken together, these data suggest that the lower levels of total testosterone in the serum of Beijing residents are due to a decrease in testosterone production rather than an increase in testosterone MCRs in these subjects. Additionally, these differences are probably due to dietary/environmental factors.
View larger version (19K):
[in this window]
[in a new window]
Figure 4. Testosterone MCRs (left axis) and production rates (right axis) in Caucasians living in the U.S. [Cau(U)] and in Chinese living in the U.S. [Ch(U)] and China [Ch(C)]. See Subjects and Methods for details of the infusions. Chinese living in China had significantly lower testosterone production rates than Chinese living in the U.S. (P = 0.034).
View larger version (22K):
[in this window]
[in a new window]
Figure 5. Plasma total testosterone and bioactive testosterone (Free T, left axis) and SHBG (right axis) in Caucasians living in the U.S. [Cau(U)] and Chinese living in either the U.S. [CH(U)] or China [Ch(U)]. + indicates outliers, as described in Fig. 2. Chinese living in China had significantly lower total testosterone and SHBG levels than Chinese living in the U.S. (P = 0.031 and 0.038, respectively).
Having detected differences in androgen metabolism between Beijing and Pennsylvania Chinese, we concluded that study of genetic differences must necessarily involve study of two groups living in the United States. Our experimental design allowed this comparison because the isotopic kinetic studies involved Chinese and Caucasian men both living in Pennsylvania. Surprisingly, we detected no significant differences in total testosterone, free and weakly bound testosterone, SHBG, and testosterone production rates between these two groups of Caucasian and Chinese men (Fig. 5). These data suggest that no major genetic differences exist with respect to these parameters.
Discussion
Top
Abstract
Introduction
Subjects and Methods
Results
Discussion
References
Epidemiological data have demonstrated a striking reduction in the incidence of clinically apparent, but not latent, prostate cancer in Chinese men living in China than in Caucasians living in western countries (1, 2, 3, 4). Extensive studies demonstrate similar findings in Japanese men (20, 21, 22, 23). Our prior studies of 3-Diol-G and 3-A-G levels suggested that genetic alterations of 5-reductase might be present in Chinese men (7, 8). We hypothesized that a resultant lowering of prostatic tissue DHT levels might explain the lack of progression from latent to clinically apparent prostate cancer in the Chinese. However, we recognized that dietary or other environmental factors altering androgen metabolism might also be important.
The present study sought to examine both genetic as well as environmental/dietary influences on androgen metabolism. Initially, we used two proven methods to examine 5-reductase activity in patients: determination of 5ß:5 ratios of excreted urinary steroids and isotopic kinetic measurements of testosterone to DHT conversion (9, 10, 11). Neither method demonstrated diminished 5-reductase activity (or even a trend in that direction) in Chinese men living in China compared with those in the United States. Taken together, these studies suggest that there are no major genetic differences in 5-reductase activity levels between Chinese and Caucasian subjects. Dietary/environmental factors did appear important. Study of one group of Chinese men living in the United States and another living in Beijing, China, allowed us to directly test these factors. Significant differences in serum testosterone levels and production rates between these two groups clearly substantiated the importance of these nongenetic influences.
Our prior data demonstrated a clear reduction in 5-reduced androgen metabolites such as 3-Diol-G and 3-A-G levels in the plasma of Chinese compared to Caucasian subjects (7, 8). The present study provides an explanation for these differences. Our isotopic as well as ketosteroid data suggest that a lowering of androgenic precursors, which serve as substrates for these metabolites, is responsible rather than a diminution of 5-reductase enzyme activity.
It should be pointed out that our original studies comparing Chinese and Caucasian subjects were predicated on the concept that plasma 3-Diol-G and 3-A-G levels primarily reflect the rate of tissue 5-reduction of androgens (7, 8). This idea originated from studies of Mahoudeau et al. (17) and was extended by Toscano and Horton (24) and others (25, 26). Later studies, however, indicated that steroids secreted by the adrenal, such as androstenedione and androsterone, can also contribute substantially to the levels of 3-Diol-G and 3-A-G (19, 27). Based upon this reasoning, we considered it necessary to examine these two components directly by determining 5-reductase activity by isotopic and urinary methods and the levels of precursor substrates by measuring ketosteroids in urine. The results point to decreased ketosteroid precursor levels as the primary factor to explain lower plasma 3-Diol-G and 3-A-G levels.
Other investigators have also demonstrated lower 17-ketosteroid levels in Oriental than in Caucasian subjects (28). Although genetic differences among groups provide one potential explanation for this finding, dietary or other environmental factors could equally explain the differences observed. Unfortunately, we did not compare ketosteroid levels in Chinese living in the United States with those living in China to address the environmental/dietary issue directly. However, our isotopic kinetic studies do shed some light on this issue. Testosterone production rates, SHBG levels, and serum total testosterone levels were substantially lower in Chinese living in Beijing than in those living in the United States. It should be noted that the physiological interactions between non-SHBG-bound testosterone and its effect on testosterone MCR and total testosterone are complex and cannot be inferred from our results. Taken together, our findings point directly to environmental/dietary influences on androgen metabolism. These factors could also have contributed to the differences in ketosteroid levels observed between Caucasians living in the United States and Chinese living in Hong Kong. Further studies of ketosteroid excretion must be now conducted in Chinese living in the United States and compared with those of Chinese living in China.
Androgen metabolism and plasma levels in patients who ultimately develop prostate cancer or currently carry this diagnosis have been the subject of several very recent studies (5, 20, 29, 30, 31, 32, 33, 34, 35, 36, 37). In one, androgen levels were compared in groups of Japanese and Dutch men (20) with the finding of lower androgen levels in Japanese men. Another failed to find lower testosterone levels in Japanese subjects living in Japan compared to Caucasians living in the United States (5). These observations, taken in context with our results, emphasize the complexity of distinguishing genetic from environmental/dietary effects. In addition, it is necessary to prove cause and effect relationships. Although lowered androgen production could provide a hormonal explanation for the different incidence rates of clinically evident prostate cancer between Japanese and Europeans, these findings may equally represent associated, but not causally related, phenomena. Further studies are clearly necessary to distinguish among the various possibilities.
Epidemiological studies provide strong evidence that environmental/dietary factors are important in accelerating the transition between latent and clinically apparent prostate cancer. Japanese men experience an increase in the incidence of clinically apparent prostate cancer when they move to Hawaii (23). African men have a much lower incidence of prostate cancer than do African-Americans. An increase in the rate of clinically apparent prostate cancer is associated with increases in fat intake. There are also other data suggesting that increased fat intake can increase the levels of androgen production (29). Further studies will be necessary to assess the influence of diet and body composition or body surface area on androgen metabolism. The present study, although providing no definitive mechanisms, suggests the importance of dietary or other environmental influences on androgen metabolism.
Several limitations are inherent in the studies reported. We recognize that the limited numbers of study subjects available and the nonparametric distribution of data reduce the power of statistically based conclusions. Nonetheless, the absence even of trends supporting a diminution of 5-reductase activity in the Chinese men strengthens the conclusions of this study. Our data do not preclude the possibility that certain tissues may have isolated alterations of either type I or type II 5-reductase activity (38, 39). Thus, it remains possible that Chinese subjects could have diminished 5-reductase activity isolated to hair follicles or prostate. Additional studies are required to address these issues. Finally, the exact dietary or environmental factors influencing androgen metabolism have not been identified in our studies.
In conclusion, dietary or environmental factors, and not a diminution of 5-reductase, appear to be responsible for differences in androgen metabolism between Caucasians living in the United States and Chinese living in China. These results highlight the necessity of carefully controlling for multiple factors before concluding that the observed differences result from genetic factors.
Introduction
Top
Abstract
Introduction
Subjects and Methods
Results
Discussion
References
CHINESE men living in the Far East develop clinically apparent prostate cancer at a rate nearly 1/10th that in Caucasian men living in Western countries (1, 2, 3). The incidence of latent prostate cancer, on the other hand, is similar in the two populations. One potential explanation for this reduction in prostatic cancer risk in Chinese men is a genetically controlled diminution in 5-reductase activity. This would result in a lowering of prostate tissue dihydrotestosterone (DHT) levels and potentially of the rate of progression from latent to clinically apparent prostate cancer (4, 5). Alternatively, environmental or dietary factors that are prevalent in China but not in western countries could provide an explanation for the differences in cancer risk (6). These possibilities have led us to a series of studies comparing Chinese with Caucasian subjects.
We initially demonstrated that Chinese men have less facial and chest hair than age-matched Caucasian men and suggested the possibility that reduced conversion of testosterone to DHT in the Chinese might explain this difference (7, 8). In addressing this issue, we detected lower levels of the 5-reduced androgen metabolites, 3-androstanediol glucuronide (3-Diol-G) and 3-androsterone glucuronide (3A-G) in plasma of normal Chinese males and females compared to their Caucasian counterparts. It was apparent to us, however, that plasma levels of these conjugated metabolites provide only an indirect measure of tissue 5-reductase activity. The amount of circulating 3-Diol-G and 3-A-G is determined not only by tissue 5-reductase activity, but also by the amount of precursor ketosteroid secreted by the gonads and adrenals (9, 10). Accordingly, the present study used more direct measurements of 5-reductase activity and quantitative assessment of ketosteroids to evaluate differences between Chinese and Caucasian subjects. We chose techniques previously used to demonstrate 5-reductase deficiency in subjects with an inherited decrease in the activity of this enzyme. These methods involved measurement of the 5ß:5 ratios of several urinary metabolites using gas chromatography and mass spectrometry as well as isotopic kinetic determinations of the conversion of testosterone to DHT (9, 11).
The data reported here suggest that a genetic alteration of 5-reductase activity does not explain the reduced 3-Diol-G and 3-A-G levels in Chinese men. Instead, reduced levels of the androgenic ketosteroid precursors of these plasma metabolites provide the most likely explanation. Dietary or other environmental factors appear to alter the serum levels and production rates of testosterone of Chinese men living in China. These factors require consideration when postulating a causal link between the reduction of prostate cancer incidence, reduction of ketosteroid excretion, and lowered testosterone production rates in Chinese men living in China.
Subjects and Methods
Top
Abstract
Introduction
Subjects and Methods
Results
Discussion
References
Comparison of urinary metabolites in Caucasian and Chinese men and women
Overnight urine samples were obtained from 20 Caucasian men and 20 women living in the United States and from 20 Chinese men and 20 women living in Hong Kong. All subjects were normal medical students taking no medications that would alter androgen levels. Thirty-eight steroids were measured by gas chromatography-mass spectrometry (12). Assessment of 5-reductase activity, as previously described (9, 10), involved calculation of the ratios of four 5ß-:5-steroid pairs measured in overnight urine samples. For assessment of ketosteroid excretion, values were normalized by dividing by the urinary creatinine level. Total ketosteroid levels were calculated by adding all individual ketosteroid values together. These are expressed per mg creatinine to correct for variances in duration of collection of samples.
Measurement of androgens and sex hormone-binding globulin (SHBG) in Caucasian and Chinese men
Serum samples were obtained from an additional 10 Caucasian men (aged 22–27 yr) and 10 Chinese men (aged 20–37 yr) living in Pennsylvania and from 10 Chinese men (aged 24–39 yr) living in Beijing, China. Subjects living in Pennsylvania were examined at Pennsylvania State University Hospital and had a normal physical examination and no evidence of underlying illness. Subjects living in Beijing were examined by one investigator (G.-y.Z.) who noted that 9 of 10 had fathered children, and 1 had normal sexual function but was using condoms for contraception. No evidence of endocrine disorders or other illness was detected. Testis size was normal in all men, ranging from 15 mL (right and left testes) to 25 mL (right and left testes). Samples obtained from these subjects were used for measurement of testosterone, bioactive testosterone, and SHBG levels. All volunteers signed informed consents, and all protocols were approved as required by the institutions where the research was conducted. Total testosterone was measured by RIA as previously described (7, 8). Bioactive testosterone (free and weakly bound) was measured using the ammonium sulfate precipitation method (13, 14). SHBG was measured using a solid phase 125I immunoradiometric assay kit from Diagnostic Products Corp. (Los Angeles, CA).
Measurement of the interconversion of androgens in Caucasian and Chinese men
Eighty microcuries of [3H]testosterone were infused iv over a 3-h period after a 5-mL loading dose bolus of 2.5 µCi was given to the 30 male subjects. Identical infusion pumps and aliquots from centrally prepared radiolabeled testosterone were used for studies in the United States and China. Sixty milliliters of blood were collected after 2, 2.5, and 3 h of infusion. Serum was obtained and stored at -20 C until processed. Samples from Chinese men were flown to Pennsylvania frozen and processed identically to those collected in Pennsylvania. Upon thawing, the volume was measured, and [14C]testosterone, DHT, and androstenedione were added to each sample to determine recoveries. This was followed by 3 extractions with 200 mL methylene chloride. The methylene chloride fractions were combined, partially evaporated, backwashed twice with 3 mL distilled water, and taken to dryness. The samples were then purified by thin layer chromatography and high pressure liquid chromatography using solvent systems previously developed (15, 16) and outlined in Fig. 1. Calculations of conversion ratios and MCRs followed the methods of Mahoudeau et al. as extensively described previously (17).
View larger version (11K):
[in this window]
[in a new window]
Figure 1. Purification system to separate androgens used in this study. TLC, Thin layer chromatography; HPLC, high pressure liquid chromatography.
Data analysis
All samples were collected in the morning to minimize differences due to circadian variation in steroids. Due to the small number of individuals in each group and the known lack of normal distribution of hormonal data in these populations (18), data were expressed as median values with 25th and 75th percentiles in parentheses (in the tables) or were presented as box plots of the 25th and 75th percentiles with the median indicated by the line within the box and the range depicted by the bars outside the box (in the figures). Testing of statistical differences among groups used Kruksal-Wallis one-way ANOVA (when three groups were compared) and Mann-Whitney tests (when two groups were compared). All analyses were performed using the Systat computer package (Systat, Evanston, IL). The Bonferonni correction was used to compensate for multiple urine comparisons, and only P < 001 was considered statistically significant.
Results
Top
Abstract
Introduction
Subjects and Methods
Results
Discussion
References
Initial studies addressed the hypothesis that Chinese subjects have lower levels of 5-reductase activity than their Caucasian counterparts. The ratios of 5ß-:5-reduced urinary steroids were determined as a method to detect these differences. No significant differences in 5ß:5 ratios (or even suggestive trends) were found to suggest a diminution of 5-reductase activity in the Hong Kong Chinese compared to the Pennsylvania Caucasian medical students. (Table 1). Levels in the Chinese subjects were 6- to 30-fold lower than those in patients with known 5-reductase deficiency (9, 10) and were essentially the same as those in Caucasians. We next used the more sensitive isotopic kinetic methodology to enhance the likelihood of detecting small differences in 5-reductase activity (10, 11). To evaluate the possibility that dietary/environmental factors might confound interpretation, we compared Chinese living in Pennsylvania with Chinese living in Beijing and with Caucasians living in Pennsylvania. The percent conversion of testosterone to DHT as a direct measure of 5-reductase activity did not differ among the three groups studied (Fig. 2). As an additional marker of androgen metabolism, the conversion of testosterone to androstenedione also did not differ among these three groups of subjects (Fig. 2). Considering both the urinary and plasma isotopic data, we concluded that the Chinese subjects, whether living in China or in the United States, did not have lower 5-reductase activity than their Caucasian counterparts.
View this table:
[in this window]
[in a new window]
Table 1. Comparison of 5ß:5 ratios of urinary metabolites
View larger version (16K):
[in this window]
[in a new window]
Figure 2. Conversion of testosterone (T) to DHT or androstenedione (4A). Cau(U), Caucasians living in the U.S. Ch(U), Chinese living in the U.S. Ch(C), Chinese living in China. + indicates values lying greater than 1.5 times the difference between the 25th and 75th percentile values above or below the box. All statistical comparisons were nonsignificant. In the text, Beijing Chinese is frequently used in place of Ch(C), and Pennsylvania Chinese is used for Ch(U).
An alternative hypothesis suggested that the lower plasma 5-reduced metabolites in Chinese subjects than in their Caucasian counterparts might reflect a reduction of androgenic ketosteroid precursors as first proposed by Rittmaster and colleagues (19). To test this possibility, we measured a series of urinary ketosteroids, which serve as markers for adrenal androgen production. We detected lower levels of several of these components in the urines of Chinese compared to Caucasian subjects (Table 2). Androsterone and etiocholanolone were significantly lower in the Chinese subjects (data from males and females combined, P < 0.001 and P < 0.001, respectively), and 11ß-hydroxyetiocholanolone was marginally lower (P = 0.025). To integrate data from multiple steroid measurements, we also calculated the total level of excreted ketosteroids in each subject. With this methodology, we found total ketosteroid excretion to be significantly lower in Chinese than in Caucasian subjects (P = 0.006; Fig. 3). When examined separately by gender, the differences appeared to be greater in females than in males, but variances were substantial. As a control for potential confounding factors, such as the amount of creatinine excreted, we also measured glucocorticoid metabolites. These levels, in contrast, were not lower in Chinese than in Caucasian subjects (Table 2). Interestingly, three minor compounds, 6ß-hydroxycortisol, 20-dihydrocortisol, and 20ß-dihydrocortisol, were higher in the Chinese group (data from males and females combined, P = 0.044, 0.02, and 0.007, respectively), but not sufficiently to reach the Bonferonni limits of significance (i.e. P < 0.001).
View this table:
[in this window]
[in a new window]
Table 2. Urinary levels of individual ketosteroid and glucocorticoid metabolites
View larger version (18K):
[in this window]
[in a new window]
Figure 3. Ratios of total urinary ketosteroids (micrograms per dL) to creatinine (milligrams per dL) in Caucasian (Cau) and Chinese (CHI) subjects. + indicates outliers, as described in Fig. 2. Chinese subjects had significantly lower ketosteroids (*, P = 0.006). Individual comparisons between Caucasian and Chinese males and Caucasian and Chines females were not statistically significant.
As a potential explanation for the lowered ketosteroid excretion in subjects living in China, we postulated that environmental, dietary, or genetic effects could either act independently or in concert to alter androgen production or metabolism. Lacking ketosteroid data in Chinese men living in Pennsylvania (Pennsylvania Chinese), we instead analyzed isotopic kinetic data to compare androgen production and metabolism directly in Pennsylvania Chinese with Beijing, Chinese. Notably, we detected a reduction in testosterone production rates (Fig. 4) in the Beijing compared with the Pennsylvania Chinese (P = 0.034). In contrast, testosterone MCRs did not differ between these two groups (Fig. 4). We also detected significantly lower levels of total testosterone and SHBG levels (P = 0.031 and 0.038, respectively) in Beijing than in Pennsylvania Chinese (Fig. 5). The lower SHBG levels probably explain the inability to demonstrate lower bioavailable testosterone levels in the Chinese men living in Beijing than in those in Pennsylvania. Taken together, these data suggest that the lower levels of total testosterone in the serum of Beijing residents are due to a decrease in testosterone production rather than an increase in testosterone MCRs in these subjects. Additionally, these differences are probably due to dietary/environmental factors.
View larger version (19K):
[in this window]
[in a new window]
Figure 4. Testosterone MCRs (left axis) and production rates (right axis) in Caucasians living in the U.S. [Cau(U)] and in Chinese living in the U.S. [Ch(U)] and China [Ch(C)]. See Subjects and Methods for details of the infusions. Chinese living in China had significantly lower testosterone production rates than Chinese living in the U.S. (P = 0.034).
View larger version (22K):
[in this window]
[in a new window]
Figure 5. Plasma total testosterone and bioactive testosterone (Free T, left axis) and SHBG (right axis) in Caucasians living in the U.S. [Cau(U)] and Chinese living in either the U.S. [CH(U)] or China [Ch(U)]. + indicates outliers, as described in Fig. 2. Chinese living in China had significantly lower total testosterone and SHBG levels than Chinese living in the U.S. (P = 0.031 and 0.038, respectively).
Having detected differences in androgen metabolism between Beijing and Pennsylvania Chinese, we concluded that study of genetic differences must necessarily involve study of two groups living in the United States. Our experimental design allowed this comparison because the isotopic kinetic studies involved Chinese and Caucasian men both living in Pennsylvania. Surprisingly, we detected no significant differences in total testosterone, free and weakly bound testosterone, SHBG, and testosterone production rates between these two groups of Caucasian and Chinese men (Fig. 5). These data suggest that no major genetic differences exist with respect to these parameters.
Discussion
Top
Abstract
Introduction
Subjects and Methods
Results
Discussion
References
Epidemiological data have demonstrated a striking reduction in the incidence of clinically apparent, but not latent, prostate cancer in Chinese men living in China than in Caucasians living in western countries (1, 2, 3, 4). Extensive studies demonstrate similar findings in Japanese men (20, 21, 22, 23). Our prior studies of 3-Diol-G and 3-A-G levels suggested that genetic alterations of 5-reductase might be present in Chinese men (7, 8). We hypothesized that a resultant lowering of prostatic tissue DHT levels might explain the lack of progression from latent to clinically apparent prostate cancer in the Chinese. However, we recognized that dietary or other environmental factors altering androgen metabolism might also be important.
The present study sought to examine both genetic as well as environmental/dietary influences on androgen metabolism. Initially, we used two proven methods to examine 5-reductase activity in patients: determination of 5ß:5 ratios of excreted urinary steroids and isotopic kinetic measurements of testosterone to DHT conversion (9, 10, 11). Neither method demonstrated diminished 5-reductase activity (or even a trend in that direction) in Chinese men living in China compared with those in the United States. Taken together, these studies suggest that there are no major genetic differences in 5-reductase activity levels between Chinese and Caucasian subjects. Dietary/environmental factors did appear important. Study of one group of Chinese men living in the United States and another living in Beijing, China, allowed us to directly test these factors. Significant differences in serum testosterone levels and production rates between these two groups clearly substantiated the importance of these nongenetic influences.
Our prior data demonstrated a clear reduction in 5-reduced androgen metabolites such as 3-Diol-G and 3-A-G levels in the plasma of Chinese compared to Caucasian subjects (7, 8). The present study provides an explanation for these differences. Our isotopic as well as ketosteroid data suggest that a lowering of androgenic precursors, which serve as substrates for these metabolites, is responsible rather than a diminution of 5-reductase enzyme activity.
It should be pointed out that our original studies comparing Chinese and Caucasian subjects were predicated on the concept that plasma 3-Diol-G and 3-A-G levels primarily reflect the rate of tissue 5-reduction of androgens (7, 8). This idea originated from studies of Mahoudeau et al. (17) and was extended by Toscano and Horton (24) and others (25, 26). Later studies, however, indicated that steroids secreted by the adrenal, such as androstenedione and androsterone, can also contribute substantially to the levels of 3-Diol-G and 3-A-G (19, 27). Based upon this reasoning, we considered it necessary to examine these two components directly by determining 5-reductase activity by isotopic and urinary methods and the levels of precursor substrates by measuring ketosteroids in urine. The results point to decreased ketosteroid precursor levels as the primary factor to explain lower plasma 3-Diol-G and 3-A-G levels.
Other investigators have also demonstrated lower 17-ketosteroid levels in Oriental than in Caucasian subjects (28). Although genetic differences among groups provide one potential explanation for this finding, dietary or other environmental factors could equally explain the differences observed. Unfortunately, we did not compare ketosteroid levels in Chinese living in the United States with those living in China to address the environmental/dietary issue directly. However, our isotopic kinetic studies do shed some light on this issue. Testosterone production rates, SHBG levels, and serum total testosterone levels were substantially lower in Chinese living in Beijing than in those living in the United States. It should be noted that the physiological interactions between non-SHBG-bound testosterone and its effect on testosterone MCR and total testosterone are complex and cannot be inferred from our results. Taken together, our findings point directly to environmental/dietary influences on androgen metabolism. These factors could also have contributed to the differences in ketosteroid levels observed between Caucasians living in the United States and Chinese living in Hong Kong. Further studies of ketosteroid excretion must be now conducted in Chinese living in the United States and compared with those of Chinese living in China.
Androgen metabolism and plasma levels in patients who ultimately develop prostate cancer or currently carry this diagnosis have been the subject of several very recent studies (5, 20, 29, 30, 31, 32, 33, 34, 35, 36, 37). In one, androgen levels were compared in groups of Japanese and Dutch men (20) with the finding of lower androgen levels in Japanese men. Another failed to find lower testosterone levels in Japanese subjects living in Japan compared to Caucasians living in the United States (5). These observations, taken in context with our results, emphasize the complexity of distinguishing genetic from environmental/dietary effects. In addition, it is necessary to prove cause and effect relationships. Although lowered androgen production could provide a hormonal explanation for the different incidence rates of clinically evident prostate cancer between Japanese and Europeans, these findings may equally represent associated, but not causally related, phenomena. Further studies are clearly necessary to distinguish among the various possibilities.
Epidemiological studies provide strong evidence that environmental/dietary factors are important in accelerating the transition between latent and clinically apparent prostate cancer. Japanese men experience an increase in the incidence of clinically apparent prostate cancer when they move to Hawaii (23). African men have a much lower incidence of prostate cancer than do African-Americans. An increase in the rate of clinically apparent prostate cancer is associated with increases in fat intake. There are also other data suggesting that increased fat intake can increase the levels of androgen production (29). Further studies will be necessary to assess the influence of diet and body composition or body surface area on androgen metabolism. The present study, although providing no definitive mechanisms, suggests the importance of dietary or other environmental influences on androgen metabolism.
Several limitations are inherent in the studies reported. We recognize that the limited numbers of study subjects available and the nonparametric distribution of data reduce the power of statistically based conclusions. Nonetheless, the absence even of trends supporting a diminution of 5-reductase activity in the Chinese men strengthens the conclusions of this study. Our data do not preclude the possibility that certain tissues may have isolated alterations of either type I or type II 5-reductase activity (38, 39). Thus, it remains possible that Chinese subjects could have diminished 5-reductase activity isolated to hair follicles or prostate. Additional studies are required to address these issues. Finally, the exact dietary or environmental factors influencing androgen metabolism have not been identified in our studies.
In conclusion, dietary or environmental factors, and not a diminution of 5-reductase, appear to be responsible for differences in androgen metabolism between Caucasians living in the United States and Chinese living in China. These results highlight the necessity of carefully controlling for multiple factors before concluding that the observed differences result from genetic factors.
