The effect of dehydroepiandrosterone on lumbar spine bone mineral density in patients with quiescent systemic lupus erythematosus




Because dehydroepiandrosterone (DHEA) is an adrenal steroid hormone with weak intrinsic androgenic properties that can be converted in peripheral tissues into more potent sex hormones, one might expect a positive effect of DHEA on bone mineral density (BMD). We evaluated the effects on lumbar BMD of oral DHEA, 200 mg/day, for 1 year in female patients with quiescent systemic lupus erythematosus (SLE).


The study subjects were 60 women with SLE. All participants gave informed consent to participate in a double-blind, placebo-controlled study on the effects of DHEA on fatigue and general well-being. BMD was measured with dual-energy x-ray absorptiometry (DEXA) at baseline and after 12 months.


Fifty-eight patients (mean age 42.6 years) could be evaluated; 2 patients (both in the DHEA group) refused to undergo DEXA a second time. In premenopausal women, DHEA did not influence BMD significantly. There was a significant increase in BMD with use of DHEA in postmenopausal women who were not receiving bisphosphonates or estrogen-containing medications. This increase was not observed in the group receiving placebo.


In premenopausal women with quiescent SLE, use of DHEA does not have a significant effect on BMD. DHEA may increase BMD in postmenopausal SLE patients if they are not already protected from bone loss by use of estrogens or bisphosphonates. Small numbers, due to the absence of stratification for menopausal status, and the use of antiresorptive agents at randomization preclude firmer conclusions based on the results of this study.

Dehydroepiandrosterone (DHEA) and its sulfated derivate DHEAS are steroid hormones that are produced by the adrenal cortex. DHEAS has weak intrinsic androgenic properties, but in peripheral tissues DHEAS is converted into testosterone, dihydrotestosterone, and estrogens (1). Serum levels of DHEAS progressively decrease with age from the third decade on (2), and a variety of age-related abnormalities, including low energy levels, osteoporosis, and immune senescence, have been linked to this decrease (3). Patients with systemic lupus erythematosus (SLE) have low levels of circulating DHEAS compared with age-matched healthy controls (4, 5), and an association between circulating DHEAS levels and bone mineral density (BMD) has been reported (6).

In SLE, oral administration of DHEA can reduce disease activity as well as the daily glucocorticoid dose needed to control disease activity (7, 8). Fatigue and low energy levels are frequent symptoms in patients with SLE, even when the disease is in remission. We performed a double-blind, placebo-controlled study in female patients with quiescent SLE to test whether administration of 200 mg/day of DHEA has a positive effect on well-being and fatigue. We report the effects of this treatment for 1 year on lumbar spine BMD.


Participants were 60 female SLE patients (9) from the departments of Rheumatology and Clinical Immunology of the University Medical Center Utrecht and the University Hospital Groningen, who consecutively gave informed consent to participate in a placebo-controlled study on the effects of administration of DHEA on fatigue and general well-being. The institutional review boards approved the study. Eligible patients were not receiving >10 mg of prednisone daily (or an equivalent dosage of glucocorticoids) at the time of inclusion or during the preceding 6 months. Both physicians and patients were instructed to keep the daily dose of glucocorticoids, calcium, vitamin D, bisphosphonates, and estrogen-containing medications constant during the 12-month study period, during which patients received a capsule containing either DHEA (200 mg) or placebo once daily. Identical capsules were provided by the hospital pharmacy. DHEA was purchased from Fagron (Nieuwerkerk aan den IJssel, The Netherlands).

Clinical evaluations were performed at study entry and at 3, 6, and 12 months. At these visits, disease activity was scored by the SLE Disease Activity Index (SLEDAI) (10). At baseline and 12 months, BMD of the lumbar spine (first through fourth lumbar vertebrae) was measured with a multiple detector fan-beam dual-energy x-ray absorptiometry (DEXA) bone densitometer (QDR Delphi; Hologic Europe NV, Zaventem, Belgium). Daily quality control was accomplished using a calibrated spine phantom to verify that the system's performance was within established parameters. Results are reported in gm/cm2. In a single patient, the minimum change between 2 scans that can be considered statistically significant (i.e., cannot be attributed to measurement error with 95% confidence) is 0.023 gm/cm2. We defined normal BMD, osteopenia, and osteoporosis according to World Health Organization criteria, with T score cutoffs set at −1.0 and −2.5. Two patients (both in the DHEA group) refused to undergo a second BMD measurement. Therefore, the study group comprised 58 patients.

Women with regular menses (n = 42) were considered premenopausal. Postmenopausal status was defined as amenorrhea for at least 1 year in women with a uterus in situ (n = 16), and in hysterectomized women by the presence of a follicle-stimulating hormone level of >35 IU/liter (n = 0). In order to evaluate the effects of DHEA on BMD, we divided the patients into 2 groups: bone protected and bone nonprotected. The bone-protected group comprised, irrespective of the use of bisphosphonates, premenopausal women (n = 42), postmenopausal women receiving estrogen-containing medications (n = 1), and postmenopausal women who were not receiving estrogen-containing medications but did use bisphosphonates (n = 6). The nonprotected group consisted of postmenopausal women who were not receiving estrogen-containing medications or bisphosphonates (n = 9).

Having ascertained that BMD scores were normally distributed, parametric statistics were applied for their evaluation. Independent-sample t-tests, the chi-square test, bivariate correlations, and univariate analysis of variance were used for statistical analysis, as appropriate. An alpha level of P < 0.05 was considered significant in all tests.


As shown in Table 1, most baseline characteristics were similar for patients receiving DHEA (n = 28) and those receiving placebo (n = 30). The only significant difference between the 2 groups was the percentage of postmenopausal patients (39% in the DHEA group and 17% in the placebo group; P = 0.02).

Table 1. Patient characteristics at study entry*
 DHEA 200 mg (n = 28)Placebo (n = 30)P
  • *

    Except where indicated otherwise, values are the number (%). DHEA = dehydroepiandrosterone; SLEDAI = Systemic Lupus Erythematosus Disease Activity Index (score ranges from 0 to 108).

  • Unpaired t-tests were used for continuous outcomes, and the chi-square test was used for proportional outcomes.

Age, mean ± SD years44.6 ± 11.940.7 ± 9.40.17
Caucasian23 (82.1)28 (93.3) 
Body mass index, mean ± SD25.0 ± 5.226.0 ± 5.70.49
Disease duration, mean ± SD years12.6 ± 7.012.3 ± 7.40.88
SLEDAI score, mean ± SD (range)2.82 ± 2.42 (0–10)3.40 ± 1.83 (0–6)0.31
Postmenopausal11 (39.3)5 (16.7)0.02
Prednisone use   
 Never used2 (7.1)3 (10.0) 
 Current user14 (50.0)17 (56.7)0.61
  Mean ± SD dose, gm/day5.71 ± 2.325.54 ± 2.000.82
Current users of calcium17 (60.7)15 (50.0)0.41
Current users of bisphosphonates6 (21.4)6 (20.0)0.89
Current users of vitamin D3 (10.7)3 (10.0)0.58
Current users of estrogens4 (14.3)7 (23.3)0.38
Osteopenia or osteoporosis of lumbar spine9 (32.1)10 (33.3)0.92

Based on their BMD at baseline, 26% of patients were osteopenic, and 7% were osteoporotic. BMD at baseline was significantly associated with postmenopausal status, current use of prednisone, age, and body mass index (data not shown). No such association was observed for disease duration and the SLEDAI score.

In 53 patients, use of prednisone, calcium, vitamin D, bisphosphonates, and estrogen-containing medications remained unchanged during the study period. One patient in the placebo group who was receiving prednisone (5 mg/day) at the start of the study stopped taking this medication after 8 months. Four patients (all in the DHEA group) required an increase in their daily dose of prednisone. In the first patient, the increase was from 7.5 mg/day to 60 mg/day, 4 months after study entry, because of an exacerbation of chronic obstructive pulmonary disease. Over 2 months the daily dose was tapered to 7.5 mg. In the second patient, prednisone was increased from 10 mg/day to 20 mg/day, 6 months after study entry, because of active skin lesions. Four months later the daily dose was tapered to 12.5 mg and remained constant for the rest of the study period. Both of these patients were receiving calcium and bisphosphonate therapy from the beginning of the study. In 2 other patients, active lupus nephritis developed 6 months after the start of the study. In both patients the daily dose of prednisone was increased to 1 mg/kg for 4 weeks, followed by tapering over 4 months to 10 mg. One of these patients began receiving oral contraceptives simultaneously with the increase in her daily prednisone dose.

At study entry, the BMD (mean ± SD) was 0.987 ± 0.144 gm/cm2 in the DHEA group and 1.009 gm/cm2 ± 0.120 in the placebo group (P = 0.53). After 12 months, these values were 0.992 gm/cm2 ± 0.147 and 1.008 gm/cm2 ± 0.122, respectively (P = 0.65). Changes in BMD were not significantly different between the 2 groups (P = 0.46) (Figure 1A).

Figure 1.

Change in lumbar bone mineral density (BMD) after use of dehydroepiandrosterone (DHEA) or placebo for 12 months. A, Total study group (n = 58). B, Effects of DHEA in relation to menopausal status. C, Effects of DHEA in relation to the presence or absence of bone protection (defined as premenopausal status and use of bisphosphonates or estrogen-containing medications). Data are shown as box plots. Each box represents the 25th to 75th percentiles. Lines outside the boxes represent the 10th and 90th percentiles. Lines inside the boxes represent the median. The broken horizontal lines reflect a change of 0.023 gm/cm2. The 4 open circles in the DHEA group represent 4 patients for whom the daily dose of prednisone was increased significantly during the study period.

A change of 0.023 gm/cm2 between 2 BMD measurements in a single patient was considered significant. Based on this definition, a significant change over 12 months was observed in 31% of patients. In the DHEA group, an increase in BMD was noted in 7 patients (25%), and a decrease was observed in 2 patients (7%). In the placebo group, 4 patients (13%) had an increase in BMD, and 5 patients (17%) had a decrease in BMD (Figure 1A).

Subgroup analyses showed that a change in BMD was dependent on menopausal status (P = 0.006) (Figure 1B). In premenopausal women, the change in BMD was not related to DHEA treatment. In contrast, all postmenopausal women who experienced an increase in BMD had received DHEA. Among postmenopausal patient, the mean change was 1.80% with DHEA and −2.32% with placebo.

Figure 1C shows a subgroup analysis based on the presence (n = 49) or absence (n = 9) of bone protection (defined as premenopausal status and use of estrogen-containing medications, bisphosphonates, or combinations of these). The effect of DHEA on BMD depended on whether or not the patient was otherwise protected for bone loss (P = 0.000). In the absence of bone protection, the mean change in BMD was 3.22% with DHEA and −5.61% with placebo. In the presence of bone protection, a change in BMD was unrelated to DHEA treatment.

The change in BMD in response to DHEA or placebo did not depend on the presence of osteopenia or osteoporosis at baseline (P = 0.59) or current prednisone treatment (P = 0.60) (data not shown). Repetition of the statistical analyses, excluding the 5 patients who had a change in their glucocorticoid dose during the study period, did not alter the results. Symptomatic vertebral fractures were not observed during the study period.

Over 12 months, the SLEDAI score (mean ± SD) changed from 2.82 ± 2.42 to 3.25 ± 2.55 in the DHEA group and from 3.40 ± 1.83 to 3.70 ± 2.51 in the placebo group. The difference between the 2 groups was not statistically significant (P = 0.79).


A robust finding of our study is that administration of DHEA, 200 mg/day for 1 year, does not affect the lumbar spine BMD in premenopausal women with quiescent SLE. Furthermore, our data suggest that DHEA may have a protective effect on bone in postmenopausal SLE patients who are not receiving estrogens or bisphosphonates (Figures 1B and C). A protective effect of oral DHEA on BMD is consistent with the observation of a small but significant increase in BMD at the hip that was observed in 14 postmenopausal women who received daily applications of a 10% DHEA cream for 12 months (11). The mean percent increase in BMD we observed with the use of DHEA for 1 year in postmenopausal women who did not receive estrogens or bisphosphonates (3.2%) is comparable with the 4% increase that was reported for use of bisphosphonates in postmenopausal women with osteoporosis (12) and in glucocorticoid-treated postmenopausal women who did not use estrogens (13).

Although our findings are consistent with our expectations, this study has some limitations. Due to the absence of stratification for menopausal status and the use of antiresorptive medication, there was a skewed distribution for the use of DHEA and placebo among the (relatively small number of) postmenopausal patients who were studied. This prevents firmer conclusions.

Van Vollenhoven et al (14) observed that administration of 200 mg of DHEA to patients with active SLE who received high-dose glucocorticoids prevented the significant reduction in BMD that occurred over a 6-month period in the placebo group; this finding suggests that DHEA may have bone-sparing effects. Our observations, however, cannot be compared with those findings, because the patients in the study by van Vollenhoven and colleagues had SLE that was more active (mean SLEDAI score 12.6 versus 3.1 in our study), more often used prednisone (100% versus 53% in our study), received a higher daily dose of prednisone (mean 46.7 mg versus 5.6 mg among users in our study), and were younger (mean age 37.3 years versus 42.7 years in our study). Furthermore, data on menopausal status and bone-protective medications were not provided. Because glucocorticoid-induced bone loss is most marked early in the course of treatment with glucocorticoids (15), it is conceivable that the bone-sparing effects of medications are most pronounced in the first months after the start of glucocorticoid treatment.

We conclude that the use of DHEA in premenopausal patients with quiescent SLE has no significant effects on lumbar BMD. In contrast, DHEA may improve BMD in postmenopausal patients with SLE, especially if they do not also use bisphosphonates or estrogens. This can be relevant for patients in whom bone-sparing therapy with bisphosphonates or estrogen-containing medications is not tolerated or is contraindicated.