Materials and Methods
Over 15 months, 71 postmenopausal women volunteered for the study in answer to an announcement in a local newspaper. Because this study was part of a larger sleep study, women complaining of periodic limb movements were not sought. None of the women spontaneously reported having periodic limb movements. Postmenopausal status was confirmed by a serum E2 level below 140 pmol/L and a serum FSH level above 30 IU/L. In two women, serum FSH levels were 28 and 29 IU/L, but they were included because of their ages (56 and 62 years). The present guidelines for safe HRT in women with intact uteri recommend a combination of progestin and estrogen. This study was designed to evaluate unopposed estrogen effects without the possible interference of progestin supplementation. For that reason, previous hysterectomy for benign indications was considered a main inclusion criterion. This study design also allowed for a double-masked trial in the absence of progestin-induced withdrawal bleeding. Many women had used ERT, and they were included in the study if they agreed to interrupt their therapies 5 months before entering the trial. Characteristics of the study subjects are listed in Table 1.
We excluded women with neurologic, severe cardiovascular, endocrinologic, or mental diseases, narcolepsy, medicated hyperlipidemia, malignancies, venous emboli, alcohol, medication abuse, or smoking (more than 10 cigarettes/day). Before accepting volunteers into the trial, we measured blood hemoglobin, leukocytes, sedimentation rate, serum thyrotrophin, free thyroxin, vitamin B12, creatinine, glucose, and cholesterol levels to exclude women with anemia, leukemia, hypothyroidism, vitamin deficiency, uremia, or diabetes, which were associated with periodic limb movements.6 The use of any medication, antioxidants, or hormones that affected the central nervous system was not permitted during the trial. Gynecologic examinations were done on all participants at the beginning of the study. All subjects signed informed consents after oral and written information was given. The study design was approved by the ethics committees of Turku University and Turku University Central Hospital.
Follow-up was 7 months and consisted of three periods with different medications: 3 months on placebo or ERT, a 1-month placebo washout period, and 3 months on ERT or placebo. Because two manufacturers were interested in participating in the study, two different transdermal E2 regimens were used. Women younger than 56 years (n = 30) were treated with gel (E2, 2.5 g/day), and those older than 55 years (n = 41) were treated with patches (estradiolhemihydrate, 50 μg/24 hours). Six person blocks were randomized in two groups according to random permuted blocks furnished by the drug manufacturers. During the study, the subjects and study personnel were not informed about the order of treatments. Those analyzing the data also were masked about the treatment order. Randomization codes were kept secret at the drug companies until completion of data analyses. For possible emergencies, individual randomization codes were kept on the research premises in sealed envelopes, which were never opened. Placebo compounds and packs were similar to estrogen in appearance. Blood samples for serum E2 levels (Spectria RIA, Orion Diagnostica, Turku, Finland) and serum FSH levels (Delfia TRIFMA, Wallac, Finland) were taken in the morning immediately before the endpoint of each treatment.
Sleep studies were done at the end of both treatments. In the sleep studies, movements were recorded with the static-charge-sensitive bed (BioMatt, Biorec Oy, Helsinki, Finland).16 The static-charge-sensitive bed is a sensitive, nondisturbing movement sensor that also enables cardiorespiratory monitoring. It detects body movements, limb movements, breathing patterns, and heart-related movements (ballistocardiogram) and is specially designed for long-term monitoring.16,17 The sensitivity of the static-charge-sensitive bed for recording limb movements is high, which makes limb electromyograms unnecessary.17 The principle of the method (Figure 1) was described elsewhere.16 Polygraphic sleep recordings included two electroencephalograms (channels C3/A2 and C4/A1), an electrooculogram, a submandibular electromyogram, and an electrocardiogram and were carried out and analyzed in 30-second periods.18 Movement arousal was defined as a movement shown by recordings from the static-charge-sensitive bed followed by alpha activity longer than 2 seconds on the electroencephalogram. The original analog signals were amplified and digitized at a frequency of 250 samples per second with 12-bit amplitude resolution and recorded with UniPlot (Unesta Oy, Turku, Finland).
Periodic limb movements were scored only if the movements were part of a series of four or more consecutive movements with a duration of 0.5–5 seconds and an interval of 5–90 seconds. Frequency of periodic limb movements was determined every hour during sleep time (PLMST index). Because movement while awake can disturb sleep, movement activity every hour of time in bed (PLMTIB index) also was measured. Frequency of periodic limb movements with cortical arousals was also calculated and expressed as frequency per hour of time in bed (ARTIB index) and per hour of sleep time (ARST index). Mean duration of movements and mean movement intervals were measured.
Women recorded their daily climacteric symptoms for 14 days before treatments (baseline) and before the end of both 3-month treatment periods before both sleep studies. Severity of the symptoms was evaluated on a six-item scale from 0 (no symptoms) to 5 (extreme symptoms) on climacteric symptoms such as hot flushes, sweating, sleeping problems, numbness, muscular pain, palpitation, headache, dizziness, anxiety, and depression. Efficacy of ERT on climacteric vasomotor symptoms, sleep problems, and headache ensured that the estrogen given was biologically active (Figure 2: values on the intensity, y, axis indicate mean symptom scores over 14 days in each condition).
We presumed that doubling of periodic limb movement frequency, from six to 12 with a standard deviation of 6 movements, was clinically significant, so the power of the present study setup was 94%. Incidences of periodic limb movements during ERT and placebo were descriptional and no statistical analyses were applied. Statistical nonparametric methods developed for crossover design of two treatments and two periods were used (Wilcoxon rank sum test).19 The carryover and the period effects were tested first. The carryover effect measured the sustained effect of ERT on the placebo night over the 1-month washout and 3 months of placebo treatment. The period effect measured whether the first study night differed from the second study night after the 4-month interval. These effects were not expected and were considered confounding factors. The significance of the carryover and period effects was P < .1. For the treatment effect P < .05 was considered statistically significant. Spearman correlation coefficients were calculated between movement variables and climacteric symptoms, serum E2 and serum FSH levels, body mass index (BMI), and age. P < .05 was considered significant. The statistics for randomization and data evaluation were computed with SAS statistical software (SAS Institute, Cary, NC).
Five subjects interrupted the treatment because of intolerable climacteric symptoms on placebo (n = 2), headache on estrogen (n = 1), fear of HRT (n = 1), and personal reasons (n = 1). Four subjects were excluded because of starting antidepressive drugs after the wash-out period (n = 1) and incomplete data (n = 3); thus, the final study group consisted of 62 women.
Incidence of periodic limb movements was evaluated by pooling the two randomizations groups together. This was possible because no carryover or period effect was observed. Median time in bed in the entire study group was 469 minutes (mean 470, range 374–539) during placebo and 474 minutes (mean 469, range 380–530) during ERT. Total median sleep times were 446 minutes (mean 444, range 309–511) and 446 minutes (mean 442, range 364–504), respectively. No subject complained of restless legs. Episodes of periodic limb movements were observed in 30 of 62 subjects (48%) during placebo and in 27 (44%) during ERT. In 17 (27%) on placebo and 19 (31%) on ERT, periodic limb movement activity exceeded the index level of 5 per hour while in bed, which is typically considered a clinically significant level. In those subjects, periodic limb movements were moderately severe; average counts of periodic limb movements per night were 37–311 for those on placebo and 38–300 for those on ERT. PLMTIB and PLMST correlated significantly (r = .99, P < .001).
We also noted that some nocturnal periodic limb movements occurred while the women were awake, especially before the initiation of sleep. When the two randomized groups were pooled together, those movements occurred while 14 women on placebo and ten on ERT lay awake. This finding was emphasized in women with very few periodic limb movements. In two women, all periodic movements (total of 11 movements while in bed) occurred when they were awake; therefore, in the entire study group, total counts of periodic limb movements per total sleep time were 0–280 for those on placebo and 0–291 for those on ERT. Median ARTIBs were 1.8 (range 0–7.3) for those on placebo and 1.4 (range 0–7.8) for those on ERT, and median ARSTs were 1.7 (range 0–7.8) and 1.3 (range 0–8.6; P = .358), respectively.
In the treatment effect analysis, subjects served as their own controls. Neither the carryover nor the period effect was seen in the variables. Time in bed and total sleep time were similar during ERT and placebo (P = .888 and .773, respectively). Estrogen replacement therapy was not superior to placebo in decreasing total periodic limb movement counts or periodic limb movement indices. Estrogen replacement therapy did not affect mean duration or mean interval of movements. Arousal indices remained similar (Table 2). In nine of 62 women, serum E2 concentrations did not go above 90 pmol/L. Excluding these women from the analysis did not change the results.
Mean serum E2 and FSH concentrations at baseline and during different treatments for the entire study group are presented in Table 3. Estrogen replacement therapy was pharmacologically active, as it effectively decreased serum FSH concentration; placebo had no effect. Although ERT effectively alleviated climacteric vasomotor symptoms and sleep complaints (Figure 2), none of the climacteric symptoms related to the incidence or occurrence of periodic limb movements. Variations in serum E2 concentrations during placebo or ERT were not explained by variations in BMI (P = .432 and .204, respectively). In contrast, serum FSH concentrations correlated with BMI. Women with higher BMIs had lower serum FSH concentrations during placebo (r = −.46, P < .001) and ERT (r = −.30, P = .018). Despite marked variations in hormone concentrations during ERT, serum E2 and FSH concentrations did not correlate with periodic limb movement variables. Age and BMI were unrelated (P = .325). Incidence, intensity, duration, and intervals of periodic limb movements were independent of age and BMI.
We found that periodic limb movements are common in postmenopausal women. Short-term ERT does not alter the incidence or intensity of periodic limb movements or the frequencies of periodic limb movement arousal indices. Although serum E2 concentration differed considerably across individuals, it did not affect periodic limb movements. Periodic limb movements also were independent of age and BMI in middle-aged and late-middle-aged women.
Estrogen might alter body movements because estrogen receptors and local estrogen concentrations were found in brain regions involved in the regulation of body movements such as the cerebral cortex, hippocampus, amygdala, thalamus, hypothalamus, basal forebrain, and the preoptic area.20 Several mechanisms have been described by which estrogen can affect neuronal function and structure. Estrogen stimulated neuronal repair and assisted neuronal survival and function through neurotrophic growth factors.21 There is also evidence that estrogen can prevent neuronal atrophy.22
Estrogen has agonistic and antagonistic effects on the dopaminergic system. Estrogen upregulates dopamine receptors11 and inteferes with catechol-O-methyl-transferases activity, the enzyme that degrades dopamine.13 Long exposure to estrogen increases dopamine uptake site density in the nigrostriatal dopaminergic system.12 Estrogen was found to promote dopamine uptake in cultures of mesencephalic neurons.23 Estrogen reduces dopamine concentration in the striatum14 and prevents overactivity of dopamine receptors.15
Estrogen action has been implied in several dopamine-dependent movement disorders. It might protect against dyskinesia because the prevalence of tardive dyskinesia after menopause was significantly higher in women than in men. Estrogen was used in the management of tardive dyskinesia.24 Estrogen suppressed levo-dopa–induced dyskinesia in parkinsonian patients, a condition attributed to the loss of dopamine terminals and postsynaptic supersensitivity.24 Neuroleptic-induced parkinsonian symptoms have been attributed to a blockade of dopamine receptors. Those and some other extra-pyramidal symptoms associated with the use of neuroleptics and are more common in women than in men. Estrogen increased the prevalence and severity of these symptoms.24
Estrogen also was linked to movement disorders associated with excessive dopamine activity such as chorea, especially during pregnancy and while using contraceptives.24 Also, periodic limb movements were found to be provoked or worsen during pregnancy.8 The hormonal condition of pregnancy consists of a high serum estrogen level accompanied by a high serum progestin level. The higher frequency of movement disorders during those states might depend more on progestin than estrogen concentration, or the estrogen effect might depend on simultaneous progestin availability. Although unopposed estrogen was found to increase the number of progesterone receptors,25 the absence of progestin in HRT might play a key role in the movement effect. In a case report of two postpartum psychoses accompanied by abnormal extrapyramidal movements, Vinogradov and Csernansky hypothesized that estrogen withdrawal after delivery might be an important factor in bringing about estrogen-induced dopamine receptor supersensitivity.26 In our study, serum E2 level had no importance: regardless of large interindividual variations in serum E2 concentrations, no association between periodic limb movements and serum concentrations was found.
The rapid effect of sex hormones on nervous tissue suggests that mechanisms other than genomic, eg, membrane effects, mainly regulate neuronal functions. The inconsistent effects of estrogen on the extrahypothalamic dopamine system might depend on the dose, duration, and timing of HRT. The average concentrations of estrogen in our study with postmenopausal ERT might have been too low to affect the incidence or intensity of periodic limb movements. Although periodic limb movements were common in our study group, they were only moderately severe. The focus of the study was on the physiologic phenomenon of periodic limb movements, a variable that was easier to measure and quantify objectively. Also, women without periodic limb movements at baseline were included to investigate whether periodic limb movements would appear with ERT. Women with systematic diseases associated with periodic limb movements, eg, anemia, leukemia, hypothyreosis, vitamin deficiency, uremia, diabetes, and narcolepsy, were effectively ruled out. Thus, periodic limb movements without subjective symptoms, as in this sample of women, might have been too mild to show any changes, and studies on women with symptomatic periodic limb movements are warranted.
The static-charge-sensitive bed method was described as appropriate for detecting periodic limb movements.17 Further, periodic limb movement indices (per hour in bed and per hour of sleep) correlated highly with each other; therefore, the all-night sleep polygraph, which disturbs a sleeper, is probably not needed when using the static-charge-sensitive bed. Advantages of the static-charge-sensitive bed method are its noninvasiveness and high sensitivity to minimal body movements.
After menopause, insomnia and sleeping problems are common complaints. Estrogen replacement therapy effectively alleviated these symptoms, resulting in better subjective sleep quality.27 Some3,28 but not other4 investigators believe that periodic limb movement is a common cause of insomnia. We had found that menopausal women with sporadic periodic limb movements did not have excessive sleep problems or insomnia compared with healthy postmenopausal women (unpublished observations). In the present study, ERT improved subjective sleep quality independently of periodic limb movements or related arousals.
Although designed as a randomized, double-masked, placebo-controlled, crossover trial, our study design had limitations and the results cannot be extrapolated to the general population without caution. We recruited women through an announcement in the newspaper, which might favor the recruitment of women with excessive sleep problems. Despite this possibility, a normal and expected variability of sleep complaints was observed. Effectively selecting out women with incipient disease that causes periodic limb movements can exclude those with a possibly beneficial estrogen effect.
After menopause, periodic limb movements are common, and clinically significant periodic limb movements (index >5) occur frequently. Unopposed ERT in doses normally used for postmenopausal HRT does not alter the incidence or intensity of nocturnal periodic limb movements. Thus, the reported beneficial effect of ERT on sleep quality after menopause27 might not be mediated through an effect on periodic limb movements. Unopposed ERT seems safe and does not provoke latent or worsen existing periodic limb movements. Whether estrogen and progestin replacement therapy has an effect remains unanswered.
1. Thorpy MJ, ed. ICSD—International classification of sleep disorders: Diagnostic and coding manual. Rochester, Minnesota: American Sleep Disoders Association, 1990.
2. Ancoli-Israel S, Kripke DF, Mason W, Kaplan OJ. Sleep apnea and periodic movements in an aging sample. J Gerontol 1985;40:419–25.
3. Zorick FJ, Roth T, Hartze KM, Piccione PM, Stepanski EJ. Evaluation and diagnosis of persistent insomnia. Am J Psych 1981;138:769–73.
4. Bixler EO, Kales A, Vela-Bueno A, Jacoby JA, Scarone S, Soldatos CR. Nocturnal myoclonus and nocturnal myoclonic activity in the normal population. Res Commun Chem Pathol Pharmacol 1982; 36:129–40.
5. Montplaisir J, Godbout R, Boghen D, DeChamplain J, Young SN, Lapierre G. Familial restless legs with periodic movements in sleep: Electrophysiologic, biochemical, and pharmacologic study. Neurology 1985;35:130–4.
6. Montplaisir J, Godbout R, Pelletier G, Warnes H. Restless legs syndrome and periodic limb movements during sleep. In: Kryger MH, Roth T, Dement WC, eds. Principles and practice of sleep medicine. 2nd ed. Philadelphia: WB Sanders, 1994:589–97.
7. Coleman RM, Pollak CP, Weitzman ED. Periodic movements in sleep (nocturnal myoclonus): Relation to sleep disorders. Ann Neurol 1980;8:416–21.
8. Nikkola E, Ekblad U, Ekholm E, Mikola H, Polo O. Sleep in multiple pregnancy: Breathing pattern, oxygenation, and periodic leg movements. Am J Obstet Gynecol 1996;174:1622–5.
9. Morgan K, ed. A research-based guide to sleep in later life. Baltimore: Johns Hopkins University Press, 1987.
10. Earley CJ, Allen RP. Pergolide and carbidopa/levodopa treatment of the restless legs syndrome and periodic leg movements in sleep in a consecutive series of patients. Sleep 1996;19:801–10.
11. Hruska RE, Silbergeld EK. Increased dopamine receptor sensitivity after estrogen treatment using the rat rotation model. Science 1980;208:1466–8.
12. Morissette M, Di Paolo T. Effect of chronic estradiol and progesterone treatments of ovariectomized rats on brain dopamine uptake sites. J Neurochem 1993;60:1876–83.
13. Ball P, Knuppen R, Haupt M, Breuer H. Interactions between estrogens and catechol amines. 3. Studies on the methylation of catechol estrogens, catechol amines and other catechols by the catechol-O
-methyltransferases of human liver. J Clin Endocrinol Metab 1972;34:736–9.
14. Foreman MM, Porter JC. Effects of catchol estrogens and catecholamines on hypothalamic and corpus striatal tyrosine hydroxylase activity. J Neurochem 1980;34:1175–8.
15. Koller WC, Weiner WJ, Klawans HJ, Nausieda PA. Influence of female sex hormones on neuroleptic-induced behavioral supersensitivity. Neuropharmacology 1980;19:387–91.
16. Polo O. Partial upper airway obstruction during sleep. Studies with the static-charge-sensitive bed (SCSB). Acta Physiol Scand 1992;145:1–118.
17. Rauhala E, Erkinjuntti M, Polo O. Detection of periodic leg movements with a static-charge-sensitive bed. J Sleep Res 1996;5:246–50.
18. Rechtschaffen A, Kales A, eds. A manual of standardized terminology, techniques and scoring systems for sleep stages of human subjects. Publication no. 204. Washington DC: National Institute of Health, 1968.
19. Jones B, Kenward MG, eds. Design and analysis of cross-over trials. London: Chapman and Hall, 1989.
20. McEwen BS, Woolley CS. Estradiol and progesterone regulate neuronal structure and synaptic connectivity in adult as well as developing brain. Exp Gerontol 1994;29:431–6.
21. Birge SJ. The role of estrogen deficiency in the aging central nervous system. In: Lobo RA, eds. Treatment of the postmenopausal woman. Basic and clinical aspects. 1st ed. New York: Raven Press, 1994:153–7.
22. Gould E, Woolley CS, Frankfurt M, McEwen BS. Gonadal steroids regulate dendritic spine density in hippocampal pyramidal cells in adulthood. J Neurosci 1990;10:1286–91.
23. Engele J, Pilgrim C, Reisert I. Sexual differentiation of mesencephalic neurons in vitro: Effects of sex and gonadal hormones. Int J Dev Neurosci 1989;7:603–11.
24. Di Paolo T. Modulation of brain dopamine transmission by sex steroids. Rev Neurosci 1994;5:27–41.
25. Brodeur P, Mockus M, McCullough R, Moore LG. Progesterone receptors and ventilatory stimulation by progestin. J Appl Physiol 1986;60:590–5.
26. Vinogradov S, Csernansky JG. Postpartum psychosis with abnormal movements: Dopamine supersensitivity unmasked by withdrawal of endogenous estrogens? J Clin Psychiatry 1990;51:365–6.
27. Polo-Kantola P, Erkkola R, Helenius H, Irjala K, Polo O. When does estrogen replacement therapy improve sleep quality? Am J Obstet Gynecol 1998;178:1002–9.
28. Guilleminault C, Raynal D, Weitzman ED, Dement WC. Sleep-related periodic myoclonus in patients complaining of insomnia. Trans Am Neurol Assoc 1975;100:19–21.