The 24-hr mean plasma concentrations of 13 hormones or hormone metabolites (cortisol, testosterone, dihydrotestosterone, dehydroisoandrosterone, dehydroisoandrosterone sulfate, androsterone, androsterone sulfate, estrone, thyroxine, triiodothyronine, LH, FSH, and prolactin) were measured in 16 rigorously screened patients (aged 55-80) with stage C or D prostate cancer and 36 normal men. Nine of the hormones showed no abnormalities in the patients but four (testosterone, dihydrotestosterone, cortisol, and estrone) showed abnormalities. Testosterone and dihydrotestosterone, which, respectively, decreased with age and showed no change with age in the normal men, rose sharply with age in the patients. The patients' curves crossed the normal curves at about age 65; patients 65 or above showed normal values while patients under age 65 showed significantly subnormal levels of both hormones: testosterone averaged 282 ng/dl in patients vs 434 ng/dl in controls (P less than 0.0001) and dihydrotestosterone averaged 70 ng/dl in patients vs 99 ng/dl in controls (P less than 0.01). Cortisol, which was age invariant in the normal men, fell sharply with age in the patients; patients under 65 had significantly elevated levels (10.1 vs 6.9 micrograms/dl; P less than 0.0001), while patients 65 or older had normal levels. Estrone levels were age invariant in both patients and controls, but the mean level in patients was markedly elevated (81 vs 47 pg/ml in controls; P less than 0.001). The cortisol/testosterone ratio almost completely separated prostate cancer patients under 65 from normal men, but did not discriminate patients 65 or older from normal. The findings indicate that prostate cancer patients under 65 differ markedly in their endogenous hormonal pattern from patients 65 or older. This leads us to propose a "two-disease" theory of prostate cancer, with possible differences in genetic factors and prognosis.

The 24-h mean plasma concentrations of androgens (dihydrotestosterone and total and free testosterone), estrogens (estrone and estradiol), and gonadotropins (LH and FSH) were measured in 35 healthy men, aged 21-85 yr, who were rigorously screened to exclude factors known or suspected to alter endocrine function. The plasma total testosterone concentration showed a slow continuous decline with age, decreasing about 35% between 21 and 85 yr of age; the free testosterone level was closely correlated with that of total testosterone over the entire observed concentration range. The concentrations of dihydrotestosterone, estrone, estradiol, and LH were age invariant. The concentration of FSH showed a continuous linear increase with age; the level at age 85 was about 2.5 times the level at age 21. The following conclusions were drawn. 1) Testosterone secretion appears to decline slowly and continuously throughout adult life in men. 2) Measurement of the plasma free testosterone level adds no independent information in healthy men, since its level is closely correlated with that of total testosterone at all concentrations. 3) The continuous rise with age in FSH concentration while LH is age invariant cannot be explained by changes in testosterone or estrogen production, but might be due to a decline of inhibin production with age.

The 24 hr mean plasma concentrations of estrone (E2) and estradiol (E2) were measured in 18 healthy, regularly cycling obese women; 16 healthy, regularly cycling nonobese women; 18 healthy obese men; and 33 healthy nonobese men. The obese men showed significant elevations of both E1 (67 pg/ml versus 49 pg/ml control; P less than 0.005) and E2 (37 pg/ml versus 28 pg/ml; P less than 0.005), but the obese women showed no significant elevation of either E1 or E2. The most likely explanation for the absence of significant hyperestrogenemia in the obese women despite evidence that such women have increased androstenedione-to-estrone conversion is that the latter source of estrogen is too small in comparison with estradiol secretion to cause a statistically detectable increment in plasma estrogen levels.

24-hour mean plasma levels of T3 and T4 were compared in 29 rigorously selected breast cancer patients (all stages) and 27 healthy women, and 8 A.M. "spot" plasma T4 levels were compared in 43 consecutive unselected breast cancer patients (all stages), 22 women with other-than-breast cancer, 21 women with miscellaneous non-cancerous illnesses, and the same 27 healthy women. The 24-hour T3 levels were the same in the breast cancer patients and healthy controls, but the 24-hour T4 levels were significantly higher in the cancer patients (7.7 vs 5.8 micrograms/dl, p less than 0.001); equal elevations were present in all stages of cancer. Spot T4 levels were likewise significantly higher in the breast cancer patients than in the healthy controls (8.8 vs 7.3 micrograms/dl, p less than 0.005). The women with other-than-breast cancer and the women with miscellaneous non-cancerous illnesses also showed significant elevations of spot T4 levels, indistinguishable from those of the breast cancer patients. It is concluded that breast cancer patients as a group show significant hyperthyroxinemia and that this finding may represent a second nonspecific abnormality of thyroid hormones in disease, hypotriiodothyroninemia (low-T3 syndrome) being the first.