Article Text


Serum prolactin stress values in patients with systemic lupus erythematosus
  1. C Dostál1,
  2. L Moszkorzová1,
  3. L Musilová1,
  4. Z Lacinová2,
  5. J Marek2,
  6. J Zvárová3
  1. 1Institute of Rheumatology, Na Slupi 4, Prague 2, 128 50 Czech Republic
  2. 23rd Medical Department, 1st Medical Faculty, Charles University, Prague 2, Czech Republic
  3. 3EuroMISE Centre, Charles University and Ac Sciences, Prague 2, Czech Republic
  1. Correspondence to:
    Professor C Dostál;

Statistics from

Over the past decade we have seen a gradual increase in reports giving more support to the hypothesis that mildly or moderately increased values of serum prolactin have a role in the pathogenesis and clinical activity of systemic lupus erythematosus (SLE). Jara et al summed up this information in an article published in a special edition of the international journal Lupus.1 “Idiopathic hyperprolactinaemia” (hyperprolactin) has been confirmed under a variety of conditions by several authors, including us,2 in 20–30% of patients with SLE investigated; nevertheless, opinion continues to vary about its connection with greatly increased clinical activity of the disease.3,4

Prolactin is a hormone with a very wide range of action, and its effect on the immune response has been proved in both animal experiments5 and in humans.6 It is also one of the stress hormones, and the physiological 24 hour curve of its serum concentrations is similar to that of the growth hormone, a near relation, but not to those of the hormones along the line connecting the hypothalamus-pituitary-adrenocortex.

In our first experiments designed to reproduce the study of hyperprolactin in patients with SLE or even rheumatoid arthritis (RA), we took account of the fact, emphasised by endocrinologists, that prolactin as a stress hormone is released in repeated pulses rather than continually; this is true at least, for the prolactin of pituitary origin, though not prolactin secreted in lymphoid paracrine tissue.7 The secretion of prolactin in the adenohypophysis is controlled indirectly by suppressing the inhibitory factor, which is synthesised and released in the hypothalamus and which is identical to dopamine.8

Stress, whether mental or physical, acts on prolactin secretion in different ways so that its serum levels rise not only in response to stress such as surgery in general anaesthesia but also after vigorous exercise and accidents, particularly after thoracic wall injury.9 For that reason, in compliance with recommendations from endocrinologists,8,10 we measured each time a number of successively taken samples. The samples were taken 2–3 hours after awakening, and in women of child bearing age 10–14 days after menses. The subjects under investigation spent the first 30 minutes relaxing physically and mentally; then a cannula was introduced into the cubital pit vein, and the first sample of blood taken after another 30 minutes, followed by two further samples each at 30 minute intervals. Intervals of 30 minutes were chosen because the biological half life of prolactin is 20–30 minutes.

Serum samples were taken from 130 subjects—that is, a total of 390 samples. These included 79 patients (73 women, six men) with a definite diagnosis of SLE. The degree of activity rated by the SLE Disease Activity Index (SLEDAI) score and specific organ involvement of the patient varied. Sixty two (78%) patients were evaluated as active—that is, the SLEDAI score was ⩾4 (mean (SD) 11.79 (10.4)). Seventeen (22%) patients were found to have inactive disease, with a SLEDAI score <4 (mean (SD) 1.06 (0.99)). The observed raised serum concentrations of prolactin did not correlate with the SLEDAI score activity (fig 1). Next, 23 patients (that is, 69 serum samples) with RA were investigated, who were also in different stages and activity of the disease. Twenty eight healthy subjects (that is, 84 serum samples), matched for age and sex with the SLE group, were investigated as controls. All patients and healthy controls receiving drugs known to influence prolactin secretion were excluded from the investigation.

The serum concentrations of prolactin were measured in duplicate by immunoradiometric assay (Immunotech, Prague). Prolactin levels gradually declined in successive serum samples (table 1). The differences in prolactin levels between samples 1 and 3 were highly significant for patients with SLE and RA. The statistical significance of this difference (between samples 1 and 3) was tested using the Wilcoxon matched pairs test in patients with SLE and RA, and healthy controls. The presence of hyperprolactin was rated as mild (200–350 mIU/l for men, 450–600 mIU/l for women), moderate (350–750 mIU/l for men, 600–1000 mIU/l for women), and high (>750 for men, >1000 for women). Serum hyperprolactin was found in 32/79 (41%) patients with SLE, of whom 20 (63%) had the mild type, nine (28%) the moderate, and three (9%) the high level type. In RA, serum hyperprolactin was found in 11/28 (39%) patients, of whom eight (73%) had the mild type, three (27%) the moderate form, and none had the high level type. In healthy controls, serum hyperprolactin was found in 5/28 (18%) subjects—all with the mild form. These figures apply only when increased concentrations of serum hyperprolactin were found in all three samples. If increased serum prolactin is found in only one or two samples the findings may easily be misinterpreted as seen from figure 1. Thus for the 28 healthy subjects investigated raised levels were found in all three serum samples in five (18%); in one or two samples in four (14%); and normal levels were found in 19 (68%). Similarly, for the 79 patients with SLE raised levels were found in all three serum samples in 29 (37%); in one or two samples in 19 (24%); and normal levels were found in 31 (39%). Finally, for the 23 patients with RA raised levels were found in all three serum samples in eight (35%); in one or two samples in 10 (44%); and normal levels were found in five (22%). Using Pearson’s χ2 test for contingency table, we determined significant difference (p<0.01) in the distribution of normal or raised serum prolactin levels when three blood samples were taken or when one or two samples only were taken, between patients with SLE/RA and healthy controls.

Our conclusions are partly in agreement with the hypothesis advanced by Jara et al.1 Firstly, the level of increased serum prolactin secreted from the pituitary gland is variable unless it is under the influence of a microadenoma, on the one hand, or unless the readings are of paracrine origin (lymphoid tissue), on the other. Secondly, the level of prolactin is highest in the first blood sample taken, declining in samples taken subsequently, which indicates a stressful situation at the time of taking the initial sample, though this is found only in patients with SLE or RA, not in healthy controls. The possible disregulation due to the presence of a systemic autoimmune disease requires further study.

Table 1

Serum prolactin values in patients with SLE, RA, and in healthy controls (in three successively taken samples)

Figure 1

Association of raised prolactin with diagnosis (number of observations). Results are shown for the three groups investigated: healthy controls, patients with SLE, and RA. (A) Percentage of subjects with increased serum prolactin levels in all three blood samples; only these subjects meet the condition of increased serum prolactin and, consequently, the presence of “idiopathic hyperprolactinaemia”. (B) Percentage of subjects with increased serum prolactin levels found in only one or two blood samples. An increase in only the first or in the first and second samples is taken as a sign of transitory reaction to stress caused by introducing a canula into the vein. (C) Percentage of subjects with normal serum prolactin levels found in all three blood samples; The between-group difference in the distribution of hyperprolactinaemia between the patients with SLE and RA and healthy controls is highly significant (p<0.009).


This study was supported by the Internal Grant Agency of Ministry of Health of the Czech Republic No NK 5370–3/99.


View Abstract

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.