Interrelations and associations of serum levels of steroids and pituitary hormones with markers of insulin resistance, inflammatory activity, and renal function in men and women aged >70 years in an 8-year longitudinal study of opposite-sex twins

Abstract

The physiological serum levels of steroids and pituitary hormones in older men and women have been sparsely reported in the literature. The aims of this study were to investigate the normal variation and sex differences in steroids and pituitary hormones in those aged >70 years, and to study the interrelation between these hormones and indicators of the metabolic syndrome, inflammatory activity, and renal function. The investigation comprised a population-based sample of pairs of white opposite-sex twins from the Swedish Twin Registry. At baseline in 1996 and at the 8-year follow-uup in 2004, serum levels of progesterone, cortisol, testosterone, estradiol, follicle-stimulating hormone, luteinizing hormone, prolactin, creatinine, C-reactive protein (CRP), and urea were analyzed. Serum levels of insulin and cystatin were analyzed only at the follow-up. The study sample included 219 men and 183 women aged 71 to 80 years (mean [SD], 74.5 [2.5] years) at baseline in 1996, and 127 men and 135 women at follow-uup in 2004. At baseline, in both men and women, the variation of progesterone in serum was positively correlated with that of estradiol (men: r = 0.226, P < 0.01; women: r = 0.115, P = NS), testosterone (men: r = 0.178, P < 0.01; women: r = 0.315, P < 0.001), and cortisol (men: r = 0.314, P < 0.001; women: r = 0.296, P < 0.001). The values of progesterone and other steroid hormones were associated with markers of insulin resistance (iie, insulin, waist circumference), inflammatory activity (ie, CRP) for progesterone (men: r = 0.267, P < 0.001; women: r = 0.150, P < 0.05), and renal function (ie, creatinine) for progesterone (men: r = 0.424, P < 0.001; women: r = 0.212, P < 0.01). Estradiol and prolactin were associated with insulin resistance, inflammation, and renal function. Furthermore, progesterone was associated with prolactin (men: r = 0.275, P < 0.001; women: r = 0.172, P < 0.05).. Among both men and women, there was a strong correlation between testosterone and estradiol (men: r = 0.753, P < 0.001; women: r = 0.526, P < 0.001); in women, there was also a link between testosterone and cortisol at follow-up (r = 0.340, P < 0.01). For progesterone, there was a significant correlation between the values of the co-twins (in 1996: r = 0.16, P < 0.05; in 2004: r = 0.45, P < 0.001). Higher serum levels of progesterone (2.0 [0.7] nmol/L in men and 1.7 [0.8] nmol/L in women) and prolactin (6 [5] microg/L in men and 8 [10] microg/L in women) were found among those who were deceased at follow-up compared with survivors (progesterone: 1.8 [0.5] nmol/L in men and 1.4 [0.6] nmol/L in women, P < 0.01; prolactin: 4 [3] microg/L in men and 5 [2] microg/L in women, P < 0.001). In this study of opposite-sex Swedish twins aged >70 years, there was a sex difference in the serum levels of steroids and pituitary hormones between men and women. Progesterone and other steroid hormones were associated with markers of insulin resistance, inflammatory activity, and renal function. Progesterone and prolactin levels were associated with increased risk of mortality in this sample.