Perturbations of hypothalamic-pituitary-gonadal axis and adrenal androgen functions in rheumatoid arthritis: an odyssey of hormonal relationships to the disease.

Rheumatoid arthritis (RA) is a heterogeneous disease with a diverse spectrum of manifestations and course of illness. Multiple factors are believed to contribute to its etiology. Nevertheless, consistent features are observed across populations, which include (1) increased familial or immunogenetic risk in younger-onset disease; (2) female predisposition, particularly during child-bearing ages; (3) predictable clinical improvement during pregnancy and worsening postpartum; and (4) increased incidence with aging, which suggest that hormonal factors influence the disease. In 1974, serum was prospectively obtained from pre-RA cases, 4 to 20 (mean = 12.0) years prior to onset of disease and concurrently from controls (CN) matched (4 CN per 1 RA) on age (+/- 2 years), race (all Caucasians), and entry menopausal status (EMS). CN have no known rheumatic disease. Pre-RA were divided into subgroups, according to EMS, i.e., premenopausal vs. non-premenopausal (peri- or post-menopausal), and either age at entry in 1974 or age at onset of RA. For example, one 3-way subgrouping includes: I. Entry premenopausal and RA onset < age 50 years; II. Entry premenopausal and RA onset age 50+ years, and III. Entry postmenopausal. The 11 youngest pre-RA (I) had a mean entry age of 29 years and RA onset of 41 years. An alternative 4-way subgrouping (a, b, c, d) divided the female subjects into premenopausal (last menstrual period [LMP], 0-31 days) and non-premenopausal major groups, as well as younger vs. older subgroups within the major EMS categories. The younger premenopausal women in each subgrouping system, that is, I or a, overlap almost entirely. Assays (RIA) of the major sex hormones were performed, e.g., luteinizing hormone (LH); follicle stimulating (FSH); estradiol (E2); progesterone (P4); and total testosterone (T); as well as adrenal hormones, including androstenendione (A4); dehydroepiandrosterone (DHEA); its sulfate (DHEAS); and cortisol (C). A significantly lower entry mean serum DHEAS level (mumol/L) was found in the pre-RA subgroup I, than in the 43 CN (2.14 +/- 0.47 vs. 3.62 +/- 0.37, respectively, (p = 0.033). The 25 older pre-RA and 100 CN (subgroups II and III) showed close mean DHEAS levels (1.89 +/- 0.30 and 1.94 +/- 0.14, respectively, p = 0.45). The serum DHEAS levels in pre-RA vs. CN subgroups were validated in a second reference laboratory. Also, the youngest pre-RA subgroup (I) showed a significant dissociation between entry serum DHEAS and cortisol levels (r = -0.660, p = 0.027), which differed (p = 0.017) from its matched CN, and from the older pre-RA (p = 0.004). Analyses of results based upon subgroupings by EMS and entry age (a, b, c, d) showed similar results. No significant differences were found between pre-RA and CN groups in levels of serum cortisol, other adrenal steroids, or the sex hormones assayed. In a sample of younger premenopausal women, significantly low serum DHEAS levels were found 4 to 20 years prior to onset of RA. Dissociation of serum cortisol and DHEAS levels was also found in the youngest, but not older, pre-RA subjects. The data suggest that subtle adrenal cortical dysfunction, manifested mainly by adrenal androgen (AA) deficiency, may either predispose to younger-onset RA or be a long-term marker in a minority subgroup of women.