- Email: [email protected]
Testicular endocrine function in the ageing male
273
Maturrtas, 7 (1985) 273-279 Elsevier
MAT 00377
Testicular endocrine function in the ageing male A. Vermeulen
and J.P. Deslypere,
Department o/ Endocrinologv, Academic Hospital, State Unrversity oj Ghent. Be&m (Received
5 April 1985; revision received 28 May 1985; accepted
28 June 1985)
The decrease of plasma testosterone (T) in elderly men may be inherent to the ageing process or secondary to environmental factors such as stress, smoking and physical activity. Plasma T. apparent free T and gonadotropin levels were determined in healthy men aged 26-90 yr living under identical conditions (as monks in a monastery). We found a highly significant age-dependent decrease in T and apparent free T levels. and an increase in gonadotropin levels. The nyctohemeral variation in T levels was lower in elderly than in younger males. When studied in a larger group, smokers had higher T levels than non-smokers, whereas in young men plasma T levels were more sensitive to stress than in elderly men. Neither moderate physical activity, diet nor residence consistently altered plasma T levels. We conclude that the age-associated decrease in plasma T levels is inherent to the ageing process and that the major cause is at the testicular level, although there is evidence for subtle alterations at the hypothalamo-pituitary level. (Key words:
Age, Testosterone,
Testicular
steroids,
Smoking,
Circadian
rhythm,
Stress)
Introduction
It has long been generally accepted that plasma testosterone (T) levels decrease in aged men. This view is based on data from an impressive number of papers from all over the world, reporting the very marked fall in testosterone concentration in the spermatic vein, the decreased responsiveness of the ageing testis to human chorionic gonadotrophin (HCG) stimulation, and histological evidence. These observations are in accordance with the age-dependent decrease in plasma androgens observed in most animals, as well as with clinical experience. Although there had always been some dissenting opinions, it is only recently that this established view has been challenged. Several authors have now suggested that this apparently age-associated decrease is not inherent to the ageing process, but is secondary to environmental factors such as minor illness and physical inactivity. The protagonists of the new theory point out that in most studies relating to the influence of age on T levels, the selection of the elderly male subjects concerned is
Correspondence ro: Professor A. Vermeulen, M.D., Afdeling Endccrinologie Academisch Ziekenhuis. Rijksuniversiteit Gent, 9000 Gent, Belgium.
0378-5122/85/.$03.30
0 1985 Elsevier Science Publishers
B.V.
en Stofwisselingsziekten,
274
biased, since the groups include persons residing in homes for the elderly or ambulatory patients attending geriatric outpatient departments. The elderly men selected are consequently not representative of the healthy elderly population as a whole, whereas the young men in the studies are generally perfectly healthy medical students or paramedical personnel. As we were among the first to report an age-dependent decrease in plasma testosterone levels, and since, in many studies, the selection of older men might indeed have been biased, we thought it worthwhile to reinvestigate this subject, paying special attention to selection criteria and studying only perfectly healthy men. We also felt it would be interesting to study systematically the influence on plasma androgen levels of a number of exogenous factors, such as smoking, exercise, diet and stress.
Methods
Testosterone (T), So-androstan-17P-ol-3-one (dihydrotestosterone DHT), apparent free testosterone concentration (AFTC), dehydroepiandrosterone (DHEA) and oestradiol (E2) were determined as previously described [20,21,23.24]. Plasma cortisol was measured by radioimmunoassay (RIA), using an antibody against cortisol-3 CMO-BSA produced in rabbits. Plasma gonadotrophins and prolactin were determined by the double-antibody technique, using IRE-kits (Fleurus, Belgium). The results are expressed in ng/ml, the reference standards being MRC 68/40 and 69/104 for luteinizing hormone (LH) and follicle-stimulating hormone (FSH), respectively, and NIH-L-S no. I for prolactin.
Results
Our aim in the first phase of the study was to select and examine a population of normal, healthy men, living under identical environmental conditions as regards times of waking and retiring, working schedules and food intake. We accordingly decided to study a group of monks (n = 71) living in a monastery. With the exception of 4 with evident symptoms of disease, all the monks living in the monastery concerned participated in the study. All participants were in good general health and were physically active. Their body weight was within 10% of ideal body weight. None were taking any medication, and neither physical examination nor routine blood tests revealed any abnormality. The plasma T and AFTC levels at 5:00 a.m. were significantly lower in healthy elderly monks (> 60 yr, mean age 71) than in young monks ( < 60 yr, mean age 37). There was also a statistically significant negative linear correlation between age and both plasma T and AFTC levels (r = -40, P < 0.01). LH levels were slighly but significantly higher in the group of old monks, pointing towards a primary testicular cause for decreased T levels in old age (Fig. 1).
AFTC
T
LH
E2
p<0.05
Young
Fig. 1. Influence AFTC. apparent
nnmks(n:33)
cl Old monks(n:38)
of age on plasma sex hormone levels in young and elderly monks. T. Testosterone; free testosterone concentration; E,, oestradiol; LH. luteinizing hormone.
FSH levels were also slightly but significantly higher in elderly monks than in young monks, the levels in the young monks all being within the normal range. When nyctohemeral variations were studied in 45 monks (22 young monks, mean age 37; 23 old monks, mean age 71) by measuring plasma T levels at 590 a.m.. 8:30 a.m., 12:OO noon, 4:00 p.m. and 7:OO p.m., it was observed that nyctohemeral variations were significantly less marked in the elderly monks than in the young monks when expressed either as an absolute value of the amplitude or as a percentage of the mean plasma level over the day. This confirmed the findings of Bremner et al. [2]. Moreover, unlike the morning values, the 790 p.m. values were not significantly higher in young than in old monks (Fig. 2). In addition, the amplitudes of the nyctohemeral variations of DHEA and cortisol were smaller in the elderly than in the young monks and, remarkably, DHEA nyctohemeral variations were found to parallel the T variations and not the cortisol variations. NYCTHEMERAL
VARIATIONS
BOO ngldl
1
a. 2114%
k
3.00 Fig. 2. Diurnal
4
8.00 variation
12.00 of plasma
1600
19.mhrr.
T levels in young and elderly
monks.
276
Although they constitute a homogeneous group as regards environmental factors, monks are obviously not representative of the general population. Accordingly, in the second phase of the study, hormonal levels were determined in a group of 300 healthy males from all social classes and occupations, living in a semi-rural suburb of a medium-sized industrial town. All were healthy ambulant males who met the same selection criteria as the monks and who volunteered to ‘cooperate in a scientific study’. Blood samples for hormone determination were taken in the morning between 8:00 and 1O:OOa.m., in the volunteer’s home surroundings rather than the outpatient departments of hospitals. We again observed an age-dependent decrease in T and AFTC levels. In an attempt to elucidate the reasons for the discrepancies between the results of similar studies reported in the literature, we next studied the influence of number of environmental factors on plasma T levels. The effect of smoking was studied in young (< 40 yr, n = 70). middle-aged (40-59 yr, n = 54) and elderly (> 60 yr, n = 57) men. All were within 10% of ideal body weight, in good general health and fairly representative of the general population of a semi-rural suburb of a medium-sized industrial town. They represented all social classes, from unskilled factory workers to high-level executives. Smokers were defined as having smoked more than 5 cigarettes a day for more than 2 yr. In all age groups, plasma T and AFTC were higher in smokers than in non-smokers, but the difference was smaller in elderly than in young subjects, the values among ex-smokers (i.e. those having stopped for at least 2 yr) lying somewhere inbetween. A significant negative linear correlation with age was again found in all three groups. LH levels, however, were similar in smokers, non-smokers and ex-smokers (Fig. 3). The effect of diet was studied by comparing plasma sex hormone levels in normal healthy subjects eating either a normal Belgian diet, a macrobiotic, or a vegetarian diet. All the subjects were non-smokers. When matched for age and body weight, the hormone values (T, AFTC, E,, LH) in each group were comparable to the values in non-smoking subjects on a normal diet (Fig. 4). Since it has been suggested that the lower T levels observed in elderly men might be related to the fact that those selected for some studies were residing in homes for the elderly, we also studied the influence of residence. Comparison of the hormone levels recorded in 48 elderly men residing in a home for the elderly with the values in another group of 44 elderly men comparable in all respects but living in their own homes, revealed no differences. It appears therefore that residence does not play any role in regard to plasma hormone levels. Since physical actiuiry might exert a determining influence on plasma T levels, we studied these levels in 19 middle-aged men (44-61 yr) who consecutively joined an extensive three-month physical training programme 6 weeks after suffering a myocardial infarction. Comparison of the values at the start of the programme with those at the end did not disclose any significant trend in plasma hormone levels, the mean levels before and after the training programme being comparable. Finally, we studied the effects of stress on T levels in young and elderly men. For this purpose, plasma T levels were measured in patients admitted to the coronary
EFFECTS OF SMOKING 1
AFTC T
ng/dl
AFTC
AFTC
T
T
AFTC SlllOk~
20 -
40-
39yrs
Fig. 3. Plasma T levels in smokers, non-smokers
00 - 79 yrr
SOyrs and ex-smokers
in three different
age groups. Abbrevia-
tions as in Fig. 1.
unit for mvocardial infarction. All were taken to this unit within 6 h after the onset of clinical signs suggestive of infarction, the diagnosis subsequently being confirmed by ECG and enzyme studies. All the subjects were smokers, and all recovered. Hormone levels were determined on the morning following their admission as well as on the 5 days following. The T levels in the group of young men (n = 14; age < 57 yr, mean 49 _t 4 (SD)) were lower than those in the group of elderly men (n = 14. age > 62 yr. mean 70 + 6
LH 3 ngIml
Macrobiotic nrl2 (33:3yrs) Norm.Eating n: 34 (26 * 4yrs)
1
T
AFTC
E2
Vo ct.monks n: 3 (6o:rgyrs) Non. mget.monks n:lS (74:EyrsJ
s
Fig. 4. Effect of diet on sex hormone levels in normal men. Abbreviations
as in Fig. 1.
278
(SD)). In young patients the levels decreased further on the second day in hospital, whereas in the elderly men the T levels remained unchanged throughout the j-day period. Plasma cortisol levels changed only minimally in both young and old subjects. The differences in the behaviour of the plasma T levels therefore cannot be explained by differences in the plasma cortisol levels. Indeed, it has been reported that stress-induced decreases in T levels may be the consequence of increased cortisol levels [7,22]. We also studied the influence of a standardised stress factor, namely insulin-induced hypoglycemia (C 40 mg/dl), on plasma T levels in 8 young (20-30 yr) and 5 elderly (> 65 yr) men who were in excellent health. Hypoglycaemia was similar in the two groups, and although peak cortisol levels were slightly higher in the elderly, their plasma T levels did not vary significantly, whereas there was a significant decrease in the young subjects. (P -z 0.05, Wilcoxon paired observations). These data on the influence of stress (myocardial infarction or hypoglycaemia) suggest that stress has less effect on T levels in elderly men than in young men. It is perhaps interesting to mention that symptoms of sympathetic stimulation such as sweating, tremors or tachycardia, were absent or much less pronounced during the insulin-induced hypoglycaemia in the elderly than in the younger group. The differences in the behaviour of the T levels might therefore be related to the reduced catecholamine secretion and resultant testicular vasoconstriction in the elderly.
Conclusion
In summary, our data show convincingly that plasma T and AFTC levels decrease with age in men, independently of environmental factors. This is in accordance with the age-dependent intratissular decrease in androgen concentration observed in our laboratory. We therefore conclude that this moderate decrease is inherent to the ageing process. However, the decrease becomes statistically significant only after the seventh decade of life. Accordingly, when the influence of age is studied, subjects over 70 yr of age should be adequately represented in the sample, which has not always been the case in past studies [1,11,15]. As a result of the reduced nyctohemeral variation in the elderly, this decrease is more pronounced during the morning hours than in the late afternoon. Moreover, T levels are more stress-sensitive in young men than in elderly men. Hence, in studies of this type, all sources of stress (such as having to go to hospital) should be avoided. Smoking appears to be accompanied by increased T levels. However, the data in the literature are divergent [3,6,8,13,16,17,19,25] and the mechanisms responsible for the increase are not known. Higher T levels in smokers are in agreement with the findings of psychological studies [14], in which smokers achieved higher male-pattern scores than non-smokers. On the other hand, diet, residence and moderate physical activity (at least when ambulant subjects are compared) do not significantly influence sex hormone levels in men.
279
References FA. Krieg M. Hormone blood levels and their interrelationships in 1 Bartsch W. Becker H, Pinkenburg normal men and men with benign prostatic hyperplasia (BPH). Acta Endocrinol (Copenh) 1979: 90: 727-736. in blood testosterone levels with 2 Bremner WJ. Vitiello MV, Prim PN. Loss of circadian rhythmicity aging in normal men. J Clin Endocrinol Metab 1983; 56: 1278-1281. 3 Briggs MH. Cigarette smoking and infertility in men. Med J Australia 1973: 63: 616-617. in human plasma. New Engl J Med 1965; 273: 902-907. 4 Coppage WS, Cooner AE. Testosterone 5 Cumming DC, Quigley ME, Yen SCC. Acute suppression of circulating testosterone levels by cortisol in men. J Clin Endocrinol Metab 1983; 57: 671-673. L. Caggiula A. The epidemiology of plasma 6 Dai WS. Kuller LH. Laporte RE. Gutai JP, Falvo-Gerard testosterone levels in middle-aged men. Am J Epidemiol 1981: 114: 804-816. of testosterone in normal adult males, J Clin 7 Doerr P, Pirke KM. Cortisol induced suppression Endocrinol Metab 1976; 43: 622-629. LE. Tuchfeld B. Plasma alcohol. smoking, hormone concentrations and 8 Dotson LE. Robertson self-reported aggression. J Stud Alcohol 1975; 36: 578-586. of testosterone and 17 KS in plasma by the double isotope 9 Candy HM. Peterson RE. Measurement dilution derivative technique. J Clin Endocrinol Metab 1968; 28: 949-956. system. In: Schneider EL. ed. The 10 Harman SAM. Clinical aspects of aging of the male reproductive aging reproductive system. New York: Raven Press. 1978: Vol. 4. 29-58. hormones in aging men. Measurement of sex steroids. 11 Harman SHM. Tsitouras PD. Reproductive basal luteininzing hormone and Leydig cell response to human choriogonadotropin. J Clin Endocrinol Metab 1980: 51: 35-40. levels and aging in males. In: Vermeulen A. Exley D. eds. 12 Kent JZ, Acone AB. Plasma testosterone Amsterdam: Excerpta Medica, Foundation International Congress Series no. 101; 1966; 31-35. U, Gyntelberg F. Coronary risk factors and plasma sex hormones. 13 Lindholm J, Winkel P, Brodthagen Am J Med 1982; 73: 648-651. analysis. New York: Pergamon Press. 14 Mausner B. Platt ES. Mausner JS. Smoking and behavioural 1971. functions in young 15 Nieschlag E, Lamers U. Freischem CW. Langer K, Wickings EJ. Reproductive fathers and grandfathers. J Chn Endocrinol Metab 1981: 55: 676-681. 16 Philips GB, Castelli WP. Abbot RD. MacNamara PM Association of hyperestrogenemia and coronary heart disease in men in the Framingham cohort. Am J Med 1983; 74: 863-869. 17 Shaarawy M, Mahmoud KZ. Endocrine profile and semen characteristics in male smokers. Fertil Steril 1982; 38: 255-257. and body build on gonadal 18 Sparrow D, Bone R. Rowe JW. The influence of age. alcohol consumption function in men. J Clin Endocrinol Metab 1980; 51: 508-512. 19 Tsitouras PD, Martin CE, Harman SM. Relationship of serum testosterone to sexual activity in healthy elderly men. J Gerontol 1982; 37: 288-293. 20 Verdonck L. Vermeulen A. Comparison of quick methods for the estimation of estradiol in plasma hy radioimmunoassay. J Steroid Biochem 1974: 5: 471-479. 21 Vermeulen A. Determination of androgens in plasma. In: James VHT. Serio M. Martini L. eds. The endocrine function of the human testis. New York; Academic Press, 1973; 91-156. 22 Vermeulen A. Testosterone in plasma: a physiopathological study. Verh K Acad Geneeskd Belg 1973; 37: 873-881. 23 Vermeulen A, Stoica T. Verdonck L. The apparent free testosterone concentration, an index of androgenicity. J Clin Endocrinol Metab 1971: 33: 759-767. 24 Vermeulen A, Verdonck L. Radioimmunoassay of 17/3-hydroxy-5a-androstane-3-one. 4-androstene3.17-dione, dehydroepiandrosterone, 17-hydroxyprogesterone and progesterone and its application to human male plasma. J Steroid Biochem 1976; 7: l-10. 25 Wintemitz WM. Quillen D. Acute hormonal response to cigarette smoking. J Clin Pharmacol 1977; 17: 389-397.
Maturrtas, 7 (1985) 273-279 Elsevier
MAT 00377
Testicular endocrine function in the ageing male A. Vermeulen
and J.P. Deslypere,
Department o/ Endocrinologv, Academic Hospital, State Unrversity oj Ghent. Be&m (Received
5 April 1985; revision received 28 May 1985; accepted
28 June 1985)
The decrease of plasma testosterone (T) in elderly men may be inherent to the ageing process or secondary to environmental factors such as stress, smoking and physical activity. Plasma T. apparent free T and gonadotropin levels were determined in healthy men aged 26-90 yr living under identical conditions (as monks in a monastery). We found a highly significant age-dependent decrease in T and apparent free T levels. and an increase in gonadotropin levels. The nyctohemeral variation in T levels was lower in elderly than in younger males. When studied in a larger group, smokers had higher T levels than non-smokers, whereas in young men plasma T levels were more sensitive to stress than in elderly men. Neither moderate physical activity, diet nor residence consistently altered plasma T levels. We conclude that the age-associated decrease in plasma T levels is inherent to the ageing process and that the major cause is at the testicular level, although there is evidence for subtle alterations at the hypothalamo-pituitary level. (Key words:
Age, Testosterone,
Testicular
steroids,
Smoking,
Circadian
rhythm,
Stress)
Introduction
It has long been generally accepted that plasma testosterone (T) levels decrease in aged men. This view is based on data from an impressive number of papers from all over the world, reporting the very marked fall in testosterone concentration in the spermatic vein, the decreased responsiveness of the ageing testis to human chorionic gonadotrophin (HCG) stimulation, and histological evidence. These observations are in accordance with the age-dependent decrease in plasma androgens observed in most animals, as well as with clinical experience. Although there had always been some dissenting opinions, it is only recently that this established view has been challenged. Several authors have now suggested that this apparently age-associated decrease is not inherent to the ageing process, but is secondary to environmental factors such as minor illness and physical inactivity. The protagonists of the new theory point out that in most studies relating to the influence of age on T levels, the selection of the elderly male subjects concerned is
Correspondence ro: Professor A. Vermeulen, M.D., Afdeling Endccrinologie Academisch Ziekenhuis. Rijksuniversiteit Gent, 9000 Gent, Belgium.
0378-5122/85/.$03.30
0 1985 Elsevier Science Publishers
B.V.
en Stofwisselingsziekten,
274
biased, since the groups include persons residing in homes for the elderly or ambulatory patients attending geriatric outpatient departments. The elderly men selected are consequently not representative of the healthy elderly population as a whole, whereas the young men in the studies are generally perfectly healthy medical students or paramedical personnel. As we were among the first to report an age-dependent decrease in plasma testosterone levels, and since, in many studies, the selection of older men might indeed have been biased, we thought it worthwhile to reinvestigate this subject, paying special attention to selection criteria and studying only perfectly healthy men. We also felt it would be interesting to study systematically the influence on plasma androgen levels of a number of exogenous factors, such as smoking, exercise, diet and stress.
Methods
Testosterone (T), So-androstan-17P-ol-3-one (dihydrotestosterone DHT), apparent free testosterone concentration (AFTC), dehydroepiandrosterone (DHEA) and oestradiol (E2) were determined as previously described [20,21,23.24]. Plasma cortisol was measured by radioimmunoassay (RIA), using an antibody against cortisol-3 CMO-BSA produced in rabbits. Plasma gonadotrophins and prolactin were determined by the double-antibody technique, using IRE-kits (Fleurus, Belgium). The results are expressed in ng/ml, the reference standards being MRC 68/40 and 69/104 for luteinizing hormone (LH) and follicle-stimulating hormone (FSH), respectively, and NIH-L-S no. I for prolactin.
Results
Our aim in the first phase of the study was to select and examine a population of normal, healthy men, living under identical environmental conditions as regards times of waking and retiring, working schedules and food intake. We accordingly decided to study a group of monks (n = 71) living in a monastery. With the exception of 4 with evident symptoms of disease, all the monks living in the monastery concerned participated in the study. All participants were in good general health and were physically active. Their body weight was within 10% of ideal body weight. None were taking any medication, and neither physical examination nor routine blood tests revealed any abnormality. The plasma T and AFTC levels at 5:00 a.m. were significantly lower in healthy elderly monks (> 60 yr, mean age 71) than in young monks ( < 60 yr, mean age 37). There was also a statistically significant negative linear correlation between age and both plasma T and AFTC levels (r = -40, P < 0.01). LH levels were slighly but significantly higher in the group of old monks, pointing towards a primary testicular cause for decreased T levels in old age (Fig. 1).
AFTC
T
LH
E2
p<0.05
Young
Fig. 1. Influence AFTC. apparent
nnmks(n:33)
cl Old monks(n:38)
of age on plasma sex hormone levels in young and elderly monks. T. Testosterone; free testosterone concentration; E,, oestradiol; LH. luteinizing hormone.
FSH levels were also slightly but significantly higher in elderly monks than in young monks, the levels in the young monks all being within the normal range. When nyctohemeral variations were studied in 45 monks (22 young monks, mean age 37; 23 old monks, mean age 71) by measuring plasma T levels at 590 a.m.. 8:30 a.m., 12:OO noon, 4:00 p.m. and 7:OO p.m., it was observed that nyctohemeral variations were significantly less marked in the elderly monks than in the young monks when expressed either as an absolute value of the amplitude or as a percentage of the mean plasma level over the day. This confirmed the findings of Bremner et al. [2]. Moreover, unlike the morning values, the 790 p.m. values were not significantly higher in young than in old monks (Fig. 2). In addition, the amplitudes of the nyctohemeral variations of DHEA and cortisol were smaller in the elderly than in the young monks and, remarkably, DHEA nyctohemeral variations were found to parallel the T variations and not the cortisol variations. NYCTHEMERAL
VARIATIONS
BOO ngldl
1
a. 2114%
k
3.00 Fig. 2. Diurnal
4
8.00 variation
12.00 of plasma
1600
19.mhrr.
T levels in young and elderly
monks.
276
Although they constitute a homogeneous group as regards environmental factors, monks are obviously not representative of the general population. Accordingly, in the second phase of the study, hormonal levels were determined in a group of 300 healthy males from all social classes and occupations, living in a semi-rural suburb of a medium-sized industrial town. All were healthy ambulant males who met the same selection criteria as the monks and who volunteered to ‘cooperate in a scientific study’. Blood samples for hormone determination were taken in the morning between 8:00 and 1O:OOa.m., in the volunteer’s home surroundings rather than the outpatient departments of hospitals. We again observed an age-dependent decrease in T and AFTC levels. In an attempt to elucidate the reasons for the discrepancies between the results of similar studies reported in the literature, we next studied the influence of number of environmental factors on plasma T levels. The effect of smoking was studied in young (< 40 yr, n = 70). middle-aged (40-59 yr, n = 54) and elderly (> 60 yr, n = 57) men. All were within 10% of ideal body weight, in good general health and fairly representative of the general population of a semi-rural suburb of a medium-sized industrial town. They represented all social classes, from unskilled factory workers to high-level executives. Smokers were defined as having smoked more than 5 cigarettes a day for more than 2 yr. In all age groups, plasma T and AFTC were higher in smokers than in non-smokers, but the difference was smaller in elderly than in young subjects, the values among ex-smokers (i.e. those having stopped for at least 2 yr) lying somewhere inbetween. A significant negative linear correlation with age was again found in all three groups. LH levels, however, were similar in smokers, non-smokers and ex-smokers (Fig. 3). The effect of diet was studied by comparing plasma sex hormone levels in normal healthy subjects eating either a normal Belgian diet, a macrobiotic, or a vegetarian diet. All the subjects were non-smokers. When matched for age and body weight, the hormone values (T, AFTC, E,, LH) in each group were comparable to the values in non-smoking subjects on a normal diet (Fig. 4). Since it has been suggested that the lower T levels observed in elderly men might be related to the fact that those selected for some studies were residing in homes for the elderly, we also studied the influence of residence. Comparison of the hormone levels recorded in 48 elderly men residing in a home for the elderly with the values in another group of 44 elderly men comparable in all respects but living in their own homes, revealed no differences. It appears therefore that residence does not play any role in regard to plasma hormone levels. Since physical actiuiry might exert a determining influence on plasma T levels, we studied these levels in 19 middle-aged men (44-61 yr) who consecutively joined an extensive three-month physical training programme 6 weeks after suffering a myocardial infarction. Comparison of the values at the start of the programme with those at the end did not disclose any significant trend in plasma hormone levels, the mean levels before and after the training programme being comparable. Finally, we studied the effects of stress on T levels in young and elderly men. For this purpose, plasma T levels were measured in patients admitted to the coronary
EFFECTS OF SMOKING 1
AFTC T
ng/dl
AFTC
AFTC
T
T
AFTC SlllOk~
20 -
40-
39yrs
Fig. 3. Plasma T levels in smokers, non-smokers
00 - 79 yrr
SOyrs and ex-smokers
in three different
age groups. Abbrevia-
tions as in Fig. 1.
unit for mvocardial infarction. All were taken to this unit within 6 h after the onset of clinical signs suggestive of infarction, the diagnosis subsequently being confirmed by ECG and enzyme studies. All the subjects were smokers, and all recovered. Hormone levels were determined on the morning following their admission as well as on the 5 days following. The T levels in the group of young men (n = 14; age < 57 yr, mean 49 _t 4 (SD)) were lower than those in the group of elderly men (n = 14. age > 62 yr. mean 70 + 6
LH 3 ngIml
Macrobiotic nrl2 (33:3yrs) Norm.Eating n: 34 (26 * 4yrs)
1
T
AFTC
E2
Vo ct.monks n: 3 (6o:rgyrs) Non. mget.monks n:lS (74:EyrsJ
s
Fig. 4. Effect of diet on sex hormone levels in normal men. Abbreviations
as in Fig. 1.
278
(SD)). In young patients the levels decreased further on the second day in hospital, whereas in the elderly men the T levels remained unchanged throughout the j-day period. Plasma cortisol levels changed only minimally in both young and old subjects. The differences in the behaviour of the plasma T levels therefore cannot be explained by differences in the plasma cortisol levels. Indeed, it has been reported that stress-induced decreases in T levels may be the consequence of increased cortisol levels [7,22]. We also studied the influence of a standardised stress factor, namely insulin-induced hypoglycemia (C 40 mg/dl), on plasma T levels in 8 young (20-30 yr) and 5 elderly (> 65 yr) men who were in excellent health. Hypoglycaemia was similar in the two groups, and although peak cortisol levels were slightly higher in the elderly, their plasma T levels did not vary significantly, whereas there was a significant decrease in the young subjects. (P -z 0.05, Wilcoxon paired observations). These data on the influence of stress (myocardial infarction or hypoglycaemia) suggest that stress has less effect on T levels in elderly men than in young men. It is perhaps interesting to mention that symptoms of sympathetic stimulation such as sweating, tremors or tachycardia, were absent or much less pronounced during the insulin-induced hypoglycaemia in the elderly than in the younger group. The differences in the behaviour of the T levels might therefore be related to the reduced catecholamine secretion and resultant testicular vasoconstriction in the elderly.
Conclusion
In summary, our data show convincingly that plasma T and AFTC levels decrease with age in men, independently of environmental factors. This is in accordance with the age-dependent intratissular decrease in androgen concentration observed in our laboratory. We therefore conclude that this moderate decrease is inherent to the ageing process. However, the decrease becomes statistically significant only after the seventh decade of life. Accordingly, when the influence of age is studied, subjects over 70 yr of age should be adequately represented in the sample, which has not always been the case in past studies [1,11,15]. As a result of the reduced nyctohemeral variation in the elderly, this decrease is more pronounced during the morning hours than in the late afternoon. Moreover, T levels are more stress-sensitive in young men than in elderly men. Hence, in studies of this type, all sources of stress (such as having to go to hospital) should be avoided. Smoking appears to be accompanied by increased T levels. However, the data in the literature are divergent [3,6,8,13,16,17,19,25] and the mechanisms responsible for the increase are not known. Higher T levels in smokers are in agreement with the findings of psychological studies [14], in which smokers achieved higher male-pattern scores than non-smokers. On the other hand, diet, residence and moderate physical activity (at least when ambulant subjects are compared) do not significantly influence sex hormone levels in men.
279
References FA. Krieg M. Hormone blood levels and their interrelationships in 1 Bartsch W. Becker H, Pinkenburg normal men and men with benign prostatic hyperplasia (BPH). Acta Endocrinol (Copenh) 1979: 90: 727-736. in blood testosterone levels with 2 Bremner WJ. Vitiello MV, Prim PN. Loss of circadian rhythmicity aging in normal men. J Clin Endocrinol Metab 1983; 56: 1278-1281. 3 Briggs MH. Cigarette smoking and infertility in men. Med J Australia 1973: 63: 616-617. in human plasma. New Engl J Med 1965; 273: 902-907. 4 Coppage WS, Cooner AE. Testosterone 5 Cumming DC, Quigley ME, Yen SCC. Acute suppression of circulating testosterone levels by cortisol in men. J Clin Endocrinol Metab 1983; 57: 671-673. L. Caggiula A. The epidemiology of plasma 6 Dai WS. Kuller LH. Laporte RE. Gutai JP, Falvo-Gerard testosterone levels in middle-aged men. Am J Epidemiol 1981: 114: 804-816. of testosterone in normal adult males, J Clin 7 Doerr P, Pirke KM. Cortisol induced suppression Endocrinol Metab 1976; 43: 622-629. LE. Tuchfeld B. Plasma alcohol. smoking, hormone concentrations and 8 Dotson LE. Robertson self-reported aggression. J Stud Alcohol 1975; 36: 578-586. of testosterone and 17 KS in plasma by the double isotope 9 Candy HM. Peterson RE. Measurement dilution derivative technique. J Clin Endocrinol Metab 1968; 28: 949-956. system. In: Schneider EL. ed. The 10 Harman SAM. Clinical aspects of aging of the male reproductive aging reproductive system. New York: Raven Press. 1978: Vol. 4. 29-58. hormones in aging men. Measurement of sex steroids. 11 Harman SHM. Tsitouras PD. Reproductive basal luteininzing hormone and Leydig cell response to human choriogonadotropin. J Clin Endocrinol Metab 1980: 51: 35-40. levels and aging in males. In: Vermeulen A. Exley D. eds. 12 Kent JZ, Acone AB. Plasma testosterone Amsterdam: Excerpta Medica, Foundation International Congress Series no. 101; 1966; 31-35. U, Gyntelberg F. Coronary risk factors and plasma sex hormones. 13 Lindholm J, Winkel P, Brodthagen Am J Med 1982; 73: 648-651. analysis. New York: Pergamon Press. 14 Mausner B. Platt ES. Mausner JS. Smoking and behavioural 1971. functions in young 15 Nieschlag E, Lamers U. Freischem CW. Langer K, Wickings EJ. Reproductive fathers and grandfathers. J Chn Endocrinol Metab 1981: 55: 676-681. 16 Philips GB, Castelli WP. Abbot RD. MacNamara PM Association of hyperestrogenemia and coronary heart disease in men in the Framingham cohort. Am J Med 1983; 74: 863-869. 17 Shaarawy M, Mahmoud KZ. Endocrine profile and semen characteristics in male smokers. Fertil Steril 1982; 38: 255-257. and body build on gonadal 18 Sparrow D, Bone R. Rowe JW. The influence of age. alcohol consumption function in men. J Clin Endocrinol Metab 1980; 51: 508-512. 19 Tsitouras PD, Martin CE, Harman SM. Relationship of serum testosterone to sexual activity in healthy elderly men. J Gerontol 1982; 37: 288-293. 20 Verdonck L. Vermeulen A. Comparison of quick methods for the estimation of estradiol in plasma hy radioimmunoassay. J Steroid Biochem 1974: 5: 471-479. 21 Vermeulen A. Determination of androgens in plasma. In: James VHT. Serio M. Martini L. eds. The endocrine function of the human testis. New York; Academic Press, 1973; 91-156. 22 Vermeulen A. Testosterone in plasma: a physiopathological study. Verh K Acad Geneeskd Belg 1973; 37: 873-881. 23 Vermeulen A, Stoica T. Verdonck L. The apparent free testosterone concentration, an index of androgenicity. J Clin Endocrinol Metab 1971: 33: 759-767. 24 Vermeulen A, Verdonck L. Radioimmunoassay of 17/3-hydroxy-5a-androstane-3-one. 4-androstene3.17-dione, dehydroepiandrosterone, 17-hydroxyprogesterone and progesterone and its application to human male plasma. J Steroid Biochem 1976; 7: l-10. 25 Wintemitz WM. Quillen D. Acute hormonal response to cigarette smoking. J Clin Pharmacol 1977; 17: 389-397.