En Español | Plain Text Version | Skip Navigation

You are here:  Home > Funding & Research > Research > Parkinson's Disease Research Web > Drug Sumaries

Estrogen Information Summary

Skip secondary menu

Join our electronic mailing list

A number of compounds were identified as candidates for further study by the Committee to Identify Neuroprotective Agents in Parkinson's (CINAPS). Of these compounds, Minocycline, Creatine , CoQ10 and GPI 1485 have been selected for testing in the Neuroprotection Clinical Trial.

ESTROGEN

Estrogen is a female sex hormone that may have neuroprotective effects for PD patients however; the mechanism of neuroprotection is unclear. The combination of estrogen with levodopa is sometimes associated with the impairment of voluntary movement (dyskinesias), and dyskinesias associated with levodopa are more frequently reported in women then in men.

Estrogen can increase the risk of uterus cancer. Side effects of estrogen include blood clots, nausea, abdominal pain, breast tenderness, irregular bleeding, headache, and hair loss. Women receiving estrogen may also experience edema, weight gain, hot flashes, mood swings and acne. Although the drug can be given easily orally, use in men is likely to cause unacceptable side effects.

The results of clinical studies performed in women indicated that estrogen replacement may have some clinical benefits for treating PD motor symptoms; however neuroprotective benefits cannot be inferred.

Scientific Rationale

The hypothesis that estrogens may be neuroprotective stems from the observation that the incidence of PD is lower in women.1-3 In addition, some women with PD have worsening of PD symptoms premenstrually.4 Use of oral contraception and increases in estrogen concentrations during pregnancy has also been associated with chorea.5,6 The combination of estrogen with levodopa is sometimes associated with dyskinesias7, and dyskinesias associated with levadopa are more frequently reported in women than in men.8,9 Mean dopa concentrations have also been reported to be lower in women.10

In preclinical evaluations, estrogen has been shown to attenuate striatal DA depletion after MPTP and 6-OHDA. Its effects may stem from altering DA uptake and possibly DA release or indirectly through actions on NO synthase (see below).11 Most animal model investigations to date have focused on DA synthesis, metabolism, reuptake, and release, as well as receptor binding and density and not on neuronal cell loss as with other reviewed medications. However, there are data that suggest multiple possible mechanisms by which estrogen may exert protective effects on nigral neurons, including antioxidant effects, inhibition of DA transporter activity and neurotrophic effects action through BDNF and related factors.10-12

  1. Am J Epidemiol 1995; 142:820-7
  2. Neurology 1999; 52:1214-20
  3. Neurology 2000; 55:1358-68
  4. Mov Disord 1986; 1:85-7
  5. Lancet 1966; 1:284-6
  6. Neurology, 1979; 29:1605-9.
  7. Lancet 1977; 2:1367-8
  8. Ann Neurol. 1996; 39:37-45
  9. Clin Neuropharmacol. 1998; 118-21
  10. Neurology 1990; 40:763-6
  11. J Neurocyt 2000:29:387-99
  12. J Lab and Clin Med 1996; 128:367-75
  13. Brain Res. 1999; 844:20-7

Animal Model Data

RODENT: Using an in vitro superinfusion technique, the effects of estrogen and its ability to prevent MPP+-induced DA release have been examined in tissue obtained from ovariectomized female rats.1 In this system, infusion with MPP induces significant rises in DA release and is believed to represent a mechanism by which this neurotoxin exerts its effects in nigrostriatal neurons. In this model, pretreatment with estrogen failed to prevent increases in nigrostriatal DA release, however concurrent infusion with MPP+ significantly reduced DA release compared to untreated MPP+- treated tissue (p<0.039). Concurrent administration resulted in release rates comparable to unlesioned tissue. These effects were dose related: only the highest dose (300 nM, not 0.3 or 3 nM) evaluated prevented DA increase (vs MPP+- group, p<0.032). Because the observed effects were immediate, this suggests a receptor mediated (i.e. non-genomic-mediated) mechanism by which estrogen exerts its effects.

The investigators proposed that, in vitro, estrogen may be protective by blocking MPP+ uptake and the resultant stimulated DA release, and that free DA is toxic to nigrostriatal neurons (i.e, free radical prevention through minimizing DA turnover). Likewise, other in vivo studies suggest that estrogen may prevent DA reuptake and restore DA turnover to normal levels through inhibition of DAT transporter. 2 This has been shown in ovariectomized mice and rats exposed to MPTP.3,4

Rapid removal of endogenous estrogen has been shown to acutely increase DA-mediated behavior, however chronic estrogen depletion diminishes such activity. This apparent paradox led to evaluation of the effects of chronic estrogen (12-weeks, estradiol valerate) replacement on spontaneous and methamphetamine-stimulated motor behavior and DA turnover in ovariectomized rats.5 Ovariectomized rats supplemented with estrogen for 12-weeks displayed spontaneous and methamphetamine-induced motor behavior comparable to unoperated control rats. In contrast, the untreated ovariectomized group's activity (spontaneous and stimulated) was significantly reduced compared to both the estrogen and control groups (p<0.05). Comparison between the groups, showed that estrogen had no effect on striatal DA turnover. However, microdialysis showed that estrogen was capable of inducing/restoring DA output induced by methamphetamine administration. In estrogen-supplemented rats, DA levels exceeded those observed in untreated ovariectomized rats (p<0.05) and in unoperated control rats. The results support that estrogen may increase DA release. The investigators hypothesized that chronic estrogen exposure induces striatal tyrosine hydroxylase activity which subsequently increases DA activity.

The stereospecificity of estradiol's neuroprotective effects were evaluated in male mice treated with MPTP.6 Mature male mice were treated with either b -estradiol, a -estradiol (2 m g/day) or vehicle for 10-days. On day 5, mice in each group were given MPTP (15 mg/kg x 4). After 10-days, mice treated with b -estradiol had no decrease in striatal DA concentrations or change in DA turnover when compared to unlesioned mice. In contrast, a -estradiol treatment was associated with significantly decreased striatal DA (vs control, p<0.01) with a tendency for a decrease in DA turn over. Neither treatment prevented loss in DAT binding following MPTP treatment, however DAT transporter mRNA levels were preserved following MPTP, regardless of treatment with either estrogen isomer. This suggests that the DAT transporter is not involved in estradiol's actions.

Estrogen's protective effects have been examined in ovariectomized female rats undergoing unilateral lesioning with 6-OHDA.7 After ovary removal, rats were either implanted with an estrogen releasing pellet (0.1 mg) or left untreated. Seven to ten days later, rats were infused with 6-OHDA into one brain hemisphere, and 7-10 days later the animals were sacrificed and the corpus striatum assessed. In the nonlesioned hemisphere, there was no difference in striatal DA concentrations. In contrast, estrogen attenuated the decrease in DA concentrations found in the untreated rats (p<0.025). Estrogen did not affect DA turnover rats in the lesioned hemisphere. These results combined with the observation that the nonlesion side in 6-OHDA treated rats did not show a significant difference between estrogen treated and untreated groups suggests that estrogen may affect DA reuptake and not its synthesis.

PRIMATE: The ability of estrogen to prevent loss in TH-positive neurons in ovariectomized African Green monkeys was assessed following short course (2 day) and long-term (30 day) estrogen replacement (150 m g/day).8 Two days of estrogen treatment prevented reduction in TH-positive neurons in SNpc in moneys ovariectomized 10 days prior. However, the loss in TH-positive neurons following prolonged (30 days) lack of estrogen exposure could not be reversed with 2 days of estrogen supplement. Interestingly, the level of TH-positive neuron loss (~35% loss) in the SNpc plateaued between the 10 and 30 days after ovariectomy in untreated rats. This led the investigators to propose that the SNpc includes a subgroup of neurons that is sensitive to estrogen's sparing effects. Continuous exposure to estrogen (30 days) in monkeys undergoing ovariectomy 30 days prior to sacrifice prevented loss of SNpc neurons.

OTHER: Estradiol pretreatment reduced neuronal injury in cultured mesencephalic neurons following exposure to glutamate and free radicals (hydrogen peroxide).9 Protection increased with increasing estradiol dose (10-100 m M). In contrast, concurrent administration (estradiol simultaneously with toxin) was not protective. Other hormones (corticosterone and testosterone) were not protective. Interestingly, neither tamoxifen (an estrogen receptor antagonist) or cycloheximide (a protein synthesis inhibitor) administration blocked estradiol's effects. This suggests that estradiol's neuroprotective actions are not mediated by the estrogen receptor and that they do not require the translation of new proteins. Additional support for an estrogen receptor-independent action stems from the observation that both b -estradiol (estrogen receptor activity) and a -estradiol (no activity and the estrogen receptor) both are protective to DA neurons. Results from cell free systems, suggest that estradiol's actions cannot be attributed to any free radical scavenging activity. The investigators hypothesized that estrogen mediates its protective actions indirectly by antagonizing glutamate-induced NO-synthase activity.

Note: Much of the preclinical data regarding estrogen's mechanism of action in neuroprotection in PD are conflicting. However there is generally consistent evidence that estrogen facilitates the maintenance of brain DA concentrations. Data are more limited to support that estrogen spares function or protects from neuronal loss.

  1. Brain Res. 1997; 764:9-16
  2. Eur J Pharmaco. 1998; 345:207-11
  3. J Neurochem. 1996; 66:658-66
  4. Brain Res. 2000; 868:95-104
  5. Brain Res. 2001; 900:163-8
  6. Synapse 2000; 37:245-51
  7. Brain Res. 1997; 767:340-44
  8. J Neurosci. 2000; 20:8604-9
  9. J Neurosci Res. 1998; 54:707-19

Pharmacokinetics (including blood brain barrier (BBB) penetration)

Estrogens cross the blood barrier, however the extent is inversely related to their degree of binding to sex-hormone binding globulin.1 Estrogen pharmacokinetics are dependent on the specific hormone and the dosage form. However, most synthetic estrogens are readily absorbed Tmax=4-10 hours, and have a T1/2=10-24 hours with repeated dosing.2, 3 Estrogens are metabolized by the liver (CYP3A4, glucuronidation) and at tissue sites. They also may undergo enterohepatic recycling which may produce secondary rises in estrogen concentrations.3 After conjugation, estrogen metabolites are renally eliminated.

  1. Am J of Physiol 1980;239:103-8.
  2. Premarin Package Insert. Wyeth-Ayerst, Laboratories, Philadelphia, PA 2001.
  3. J Pharm Pharmacol 1998;50:857-64

Safety/Tolerability in Humans

The manufacturer provides the following warning, "Estrogen can increase the risk of cancer of the uterus. Your doctor may prescribe a progestin along with estrogen to reduce this risk. Women who have had a hysterectomy do not have this risk. Side effects of estrogen include blood clots, nausea, abdominal pain, breast tenderness, irregular bleeding, headache, and hair loss." Women receiving estrogens may also experience edema, weight gain, hot flashes, mood swings and acne. Estrogens may be prothrombotic in certain patients.

  1. Premarin Package Insert. Wyeth-Ayerst, Laboratories, Philadelphia, PA 2001

Drug Interaction Potential

Because estrogens are extensively metabolized, they can be affected by hepatic enzyme-inducing medications, particularly those that alter elimination through CYP3A4 and UGT enzyme systems.1 Because endogenous microflora may contribute to the enterohepatic recycling of estrogens, there is the potential for antibiotics to decrease the effects of estrogens.2 Estrogen administered to postmenopausal women appears to have minimal potential to interact with other medications.3, 4

  1. Ann Rev Pharmacol Toxicol. 1999;39:1-17.
  2. Clin Pharmacokin 1999;36:309-13.
  3. J Clin Pharmacol 1988;28:463-6.
  4. Clin Pharmacol Ther 2000;68:412-7.

Clinical Trial/Epidemiological Evidence in Human PD

As part of a New York-based population survey, the association between PD with and without dementia and estrogen replacement therapy (ERT) was evaluated.1 Three groups of women were identified for research purposes (PD-only, n=87; PD with dementia, n=80; and normal age-matched controls, n=989) and risk factors compared. Logistic model analysis showed that after adjustment for age, education and ethnicity, ERT was associated with a significantly lower risk of dementia in women with PD (comparison of PD vs. PD with dementia; OR=0.22, 95% CI=0.05-1.0, p<0.05). Comparing women with PD and dementia to normal age-matched controls, showed similarly (OR=0.29, 95%CI= 0.05-1.0, p<0.05) that ERT use was associated with a lower risk of dementia in women with pre-existing PD. However, there was no support that ERT prevented PD from developing PD (OR=1.02, 95%CI= 0.56-1.8, p=0.9).

The effect of estrogen replacement on PD development and symptom severity was examined in women with idiopathic PD (onset < 5 years) not receiving levodopa therapy.2 All cases treated at a New York Movement Disorder Center between 1984 and 1987 were evaluated for hormone replacement (estrogen or progesterone), dopamine agonist use, age and age of symptom onset, UPDRS scores, educational level and smoking history. Among the 1168 women treated at the center, 138 met the study criteria (34 received hormone treatment, and 104 no hormone therapy). Eleven of the women received unopposed estrogen, 20 used a combination product (estrogen-progesterone) and in three the formulation was unknown. The usual dose of estrogen was 0.625 mg/day if in tablets and 0.05 mg/estradiol per day if delivered via a patch. The mean duration of estrogen use was 2.5 years prior to PD symptom onset, however the observed range was wide (15 years before and 4 years after the onset of PD). Women receiving estrogen were more likely to be receiving a dopamine agonist (12 vs. 2%) and were more educated than those who did not receive estrogen (15.5 vs 13.6 years). The average age of PD onset was lower in patients receiving estrogen therapy (by 5.6 years, p=0.004). The mean UPDS score (scales I-III) was significantly better in the estrogen group (estrogen 18.5 ± 10.8, no estrogen 27.0 ± 15.1, p<0.05). The positive effects of estrogen on UPDS were still observed when deprenyl use, educational level, smoking were considered in multivariate analysis. Overall, estrogen use was negatively correlated with UPDS score. These findings suggest estrogen use can decrease PD symptoms, however the question as to whether estrogen delays the onset of PD could not be settled.

The effects of hormonal variation and PD symptoms were evaluated in 10 women (42.3 ± 5.8 years, Hoehn and Yahr stage 1 to 3 in the off state) with idiopathic PD and normal menstrual cycles.3 These women received an average of 1.7 ± 1.1 anti-PD medications and had been experiencing symptoms for 5.8 ± 3.3 years. During the study, women were assessed at weekly visits for 5-weeks (i.e., through one menstrual cycle) in the morning in the "off-state" before medications had been taken. PD symptoms (UPDS I-III) were evaluated, as were plasma hormone concentrations. During the course of the study, mean estrogen concentration changed 137.1 ± 56.6 pg/mL and the mean change in progesterone concentration varied by 15.6 ± 9.1 ng/mL. The severity of motor symptoms fluctuated over the course of the month (UPDS motor 10.9 ± 6.4) as were activities of daily living and behavior and mood (magnitude not given). The observed changes in objective and subjective measures failed to correlate with plasma estrogen and /or progesterone concentrations. Interpretations of these observations are limited because of the women's age, the small sample size, duration of follow-up, and variation in the medications used during the course of the study.

A double-blind, placebo-controlled crossover study evaluating the effects of transdermal 17b -estradiol on PD-associated motor symptoms following levodopa administration was evaluated in 8 postmenopausal women.4 To be included, the women had to have mild to moderate PD (Hoehn and Yahr stages II-III), and to be removed from ERT, levodopa therapy (for 12-hours prior to assessments), and other anti-PD medications (adhered to by all but one patient who could not be withdrawn from selegiline) for a period of a sufficient duration as not to influence the results of the study. The study consisted of 2-two week phases (estradiol 0.1 mg/day or placebo) separated by a 2-week washout period. After 10-days of either treatment, patients were given an IV dose of levadopa with carbidopa (100 mg) and motor symptoms were assessed. In addition, the threshold dosage of levodopa (dosage at which motor symptoms improved ³ 30% over baseline) and duration of response to levodopa therapy was determined. The mean age of the women was 59 ± 4 years (mean age at menopause 48 ± 1 years) with the mean duration of PD 10 ± 1 years. Prior to study the patients' mean dosage of levodopa was 631 ± 73 mg. Ten days of estradiol therapy was able to decrease the levodopa-threshold dosage (estradiol-29 ± 1 vs placebo-29 ± 4, p £ 0.02). However, the duration and magnitude of improvement in motor symptoms did not change significantly with estradiol therapy. Estradiol also decreased the levodopa dosage required to induce dyskinesias (estradiol-23 ± 5 mg vs placebo-29 ± 5 mg, p <0.17), however this value did not reach statistical significance. Throughout the study, estrogen also failed to improve "on-time." During estradiol treatment, plasma estrogen increased from 14 ± 1 pg/mL to 325 ± 46 pg/mL. Adverse effects reported by the women receiving estradiol included breast tenderness (88%), increased dyskinesias (38%, severe in two patients), enlargement of breast cysts, perineal swelling, weight gain, bloating, ankle edema, skin rash (in one each). These results do not support a consistent positive symptomatic benefit from estrogen therapy in PD. Another similar placebo-controlled study (8-week, n=12) also found no significant benefit of estrogen supplementation on PD motor symptoms.5

The effects of estrogen therapy were determined retrospectively in nursing home patients with PD using data collected under the Medicare SAGE database between 1992 to 1996.6 In this study, 195 women with PD and who were receiving estrogen were identified and compared to 9950 women with PD not receiving estrogen. Data were gathered from quarterly nursing assessments using standard data collection forms evaluating activities of daily living, cognitive/mental status and mobility. Estrogen use was associated with better cognitive function (OR=0.475; 95%CI 0.313-0.72) and increased level of daily functioning (OR=0.35; 95%CI 0.25-0.49). With the exception of depression, which occurred more frequently in estrogen-treated women (38 vs 27%. p=0.001), there was no significant difference in psychological/behavioral symptoms between estrogen treated and untreated groups. These observations support that estrogen use may maintain functionality in women with PD. However, its retrospective, observational nature and use of non-PD specific evaluation tools limit the implications for putative neuroprotective effects from estrogen therapy.

Estrogen's effects on motor symptoms in postmenopausal women (last menses £ 2 years prior to study) with idiopathic PD was assessed in a randomized, placebo-controlled study.7 For inclusion, participants had to be receiving a stable levadopa regime and not be receiving a COMT inhibitor or hormone replacement for ³ 2 months prior to enrollment. In addition, during the week prior to study, patients had to have at least 2 "off periods" and 2-"intermediate periods" (defined as a level of functioning between the patient's usual "on-" and "off-" states) daily. Eligible patients were randomized to estrogen 0.625 mg/day (n=20) or placebo (n=20) given for a period of 8 weeks. During the last 3 days of study, patients were required to complete a diary that assessed the amount of time the individual sent in the "on-", "off-" and "intermediate-" state. This was compared to a similar diary completed 3-days prior to treatment randomization. UPDS scores were also compared at baseline and after 8-weeks of treatment. Mean age, duration of PD, time since menopause, and medication use were comparable between the 2 groups. After 8-weeks, estrogen use was associated with an increase in "on-time" (7% with estrogen vs. 0.5% decrease with placebo, p=0.001) and a decrease in "off-time" (4% with estrogen vs. 0% with placebo, p<0.004). UPDS motor scores also improved in estrogen-treated women (estrogen 3.5 points vs 0.4 points placebo, p=0.004). Other UPDS subscales failed to show any difference when the two groups were compared, and there was no significant difference in the frequency of dyskinesias. Seventy percent of patients receiving estrogen and 40% of women receiving placebo reported an adverse effect during the study. Facial flushing, lower abdominal discomfort, headache and visual blurring occurred more frequent with estrogen than placebo, but none of the side effects necessitate withdrawal from study. This study suggests that estrogen replacement may have some clinical benefit for treating PD motor symptoms however neuroprotective benefits cannot be inferred.

A population-based case control study in Olmstead County, Minnestota examined the association between PD with type of menopause (natural or surgical), age at menopause, and postmenopausal estrogen use.8 Seventy-two incident cases (1976-1995) were identified through the medical linkage system of the Rochester Epidemiology Project. Diagnoses were confirmed by a neurologist's review of the medical records and application of diagnostic criteria for PD including the presence of cardinal PD symptoms and their responsiveness to dopaminergic agents, and absence of causes of secondary Parkinsonism. General population controls were age-matched. Exposure data were also obtained through chart review. PD cases were more likely to have had a hysterectomy, with or without unilateral oophorectomy (OR 3.36, 95%CI= 0.12-1.85). The association between hysterectomy with or without unilateral oophorectomy and PD was statistically significant. The proportion of individuals undergoing bilateral oophorectomy was identical. While medical indications for hysterectomies are unclear, the investigators suggest that hysterectomy may have been related to ovarian dysfunction and hypothesize a relationship between ovarian dysfunction and PD.

  1. Neurology 1998;50:1141-43.
  2. Neurology 1999;52: 1417-21.
  3. Neurology 2000;55:1572-4.
  4. Neurology 1999;53:91-5.
  5. Clin Neuropharmacol 1999;22:93-7.
  6. Mov Disord. 2000; 15: 1119-24.
  7. Neurology 2000;54:2292-8.
  8. Mov Disord. 2001;16:830-7.

Last updated February 09, 2005