- Email: [email protected]
Risk factors and prevalence rate of restless legs syndrome among pregnant women in Taiwan
Sleep Medicine 13 (2012) 1153–1157
Contents lists available at SciVerse ScienceDirect
Sleep Medicine journal homepage: www.elsevier.com/locate/sleep
Original Article
Risk factors and prevalence rate of restless legs syndrome among pregnant women in Taiwan Pei-Hao Chen a,c,d, Kuang-Chung Liou a, Chie-Pein Chen b, Shih-Jung Cheng a,d,⇑ a
Department of Neurology, Mackay Memorial Hospital, Taipei, Taiwan Division of High Risk Pregnancy, Mackay Memorial Hospital, Taipei, Taiwan c Graduate Institute of Mechanical and Electrical Engineering, National Taipei University of Technology, Taipei, Taiwan d Mackay Medicine, Nursing and Management College, Taipei, Taiwan b
a r t i c l e
i n f o
Article history: Received 10 February 2012 Received in revised form 8 June 2012 Accepted 15 June 2012 Available online 31 July 2012 Keywords: Anemia Coffee Peptic ulcer disease Pregnancy Restless legs syndrome Sleep disorders
a b s t r a c t Background: The goal of this study was to assess the prevalence and clinical correlates of restless legs syndrome (RLS) among pregnant Taiwanese women. Methods: We enrolled 461 pregnant women (18–45 years) admitted at Mackay Memorial Hospital for delivery from September 2010 to May 2011. The face-to-face questionnaire used to gather data included assessment of RLS diagnostic criteria, and questions related to RLS. Results: The overall prevalence rate of RLS among the study participants was 10.4%; 2.8% were categorized as having chronic RLS. Participants without RLS reported higher folate and iron supplement consumption than those with RLS. Multivariate analysis revealed significant associations of RLS with anemia and peptic ulcer disease. Participants with transient RLS during pregnancy reported more regular coffee consumption before pregnancy, and better sleep latency, duration, and efficiency, than those with chronic RLS. Overall, 81.2% of RLS sufferers reported sleep disturbances. Conclusions: Our study revealed highly prevalent but poorly recognized RLS among Taiwanese pregnant women. The identification of predictors such as medical comorbidities, and protectors such as folate and iron supplements, is warranted for obstetric RLS. In most cases, symptoms began during the second or third trimester and resolved within a week after delivery. Restricted coffee consumption before pregnancy is encouraged, but further evidence is needed to support this recommendation. Ó 2012 Elsevier B.V. All rights reserved.
1. Introduction Restless legs syndrome (RLS), first given its name in 1945 by the Swedish neurologist Karl-Axel Ekbom [1], is a sleep and movement disorder characterized by unpleasant sensations in the legs occurring at rest and at night, with an irresistible urge to move. The diagnosis of RLS was upheld by the International Restless Legs Syndrome Study Group (IRLSSG), with four cardinal features based on the history of the patient [2]. It is one of the leading known causes of insomnia and common movement disorders. RLS affects 5–15% of the general population and it has an increased prevalence in older individuals [3,4]. The prevalence of RLS in Taiwanese adults was found to be 1.57%, suggesting substantially lower rates in Asian populations compared with Caucasians [5]. Collectively, findings to date offer evidence for an association between RLS and medical comorbidities such as osteoarthritis, varicose veins, type 2 diabe⇑ Corresponding author. Address: Department of Neurology, Mackay Memorial Hospital, No. 92, Sec. 2, Zhongshan N. Rd., Zhongshan Dist., Taipei City 10449, Taiwan. Tel.: +886 225433535x2316. E-mail address: [email protected] (S.-J. Cheng). 1389-9457/$ - see front matter Ó 2012 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.sleep.2012.06.008
tes, hypertension, hypothyroidism, fibromyalgia, rheumatoid arthritis, emphysema, chronic alcoholism, sleep apnea, chronic headaches, and injuries or diseases that affect the central nervous system [6]. However, the literature in this area is sparse and less consistent, so the possible relationships between these factors and RLS are not clearly defined. The development of RLS during pregnancy has been long known; Ekbom noted in his original report that 11.3% of 486 pregnant women had RLS [7]. Thus, the prevalence of RLS among pregnant women was twice that reported by Ekbom among a sample of 230 healthy women (5.7%). Studies found that RLS prevalence rates in pregnant women are similar in Western and Eastern countries [3]. To the best of our knowledge, overall reported prevalence rates of RLS in pregnant women based on IRLSSG criteria with face-toface interviews range from 10.5% in Turkey to 30% in Pakistan [8,9]. The main causes of association between RLS and pregnancy are not well known. Hypotheses for the higher prevalence of RLS during pregnancy involve reduced serum iron or folate levels and hormonal changes such as estrogen, progesterone, prolactin, and thyroid hormones [10–13]. However, previous studies have shown
1154
P.-H. Chen et al. / Sleep Medicine 13 (2012) 1153–1157
inconsistent results with respect to their close relationships [13– 17]. To date, there have been no reports on the prevalence of RLS among Taiwanese pregnant women. In this study, the purposes are 2-fold: (1) to determine the prevalence of RLS in Taiwanese pregnant women, and (2) to explore the clinical correlates of RLS in pregnancy and its impact on sleep.
2. Methods We prospectively enrolled pregnant women aged P18 years who were admitted to the Department of Obstetrics of Mackay Memorial Hospital for delivery from September 2010 to May 2011. Women were not included in the study if they had eclampsia, preeclampsia, or severe ankle edema. Specific diseases known to be causes of symptomatic RLS were not excluded. This study was approved by the Mackay Memorial Hospital Institutional Review Board and informed consent was obtained from each participant. A questionnaire was administered during the face-to-face interview. This questionnaire consisted of 20 questions, including demographic characteristics (age, height, body weight, education, race, number of previous pregnancies, number of children, etc.), personal behavior (exercise, smoking, drinking, caffeine, iron and vitamin supplement consumption), and past medical illnesses. Medical information was also obtained from the medical records of all participants, including demographic characteristics, previous pregnancies, drug history, and sleep disorders. Diagnosis of RLS was established on the basis of all four of the IRLSSG criteria. All participants were evaluated by the same neurologist (Chen P.H.), who reviewed all the completed questionnaires and interviewed the women either by phone or face-to-face if they responded positively to at least one of the four screening questions based on the IRLSSG criteria. There is a possibility of confusion between the symptoms of RLS and those of positional discomfort, hypnagogic jerks, nocturnal leg cramps, or musculoskeletal disorders. To prevent this confusion, we utilized the validated Hopkins telephone diagnostic interview [18] that contained 11 specific questions to confirm RLS and to eliminate potential false positives. Participants who fulfilled the IRLSSG criteria were asked to fill in a form that included questions to assess their RLS history and the RLS history of their first-degree relatives, as well as questions based on the John Hopkins RLS Severity Rating Scale [19] and the Chinese version of Pittsburgh Sleep Quality Index (PSQI) [20]. The John Hopkins RLS Severity Rating Scale consists of one question regarding the usual (>50% of days) time of the day that symptoms started. Severity of RLS can be classified as minimal (score 0), mild (score 1), moderate (score 2), and severe (score 3). This crude severity rating scale rates most people with RLS fairly well, but it does not take into account the intensity of the RLS symptoms and their effect on daytime functioning. The Pittsburgh Sleep Quality Index measures components of sleep quality, sleep latency, sleep duration, habitual sleep efficiency, sleep disturbances, use of sleeping medications, daytime dysfunction, and a global score. Scores range from 0 to 21, with scores greater than five indicating clinically significant sleep disturbance. Hemoglobin levels were obtained from routine blood tests between week 24 and week 28 of pregnancy. Women were considered anemic if hemoglobin levels were below 11 g/dL [21]. All participants affected by RLS underwent follow-up telephone interviews conducted by the same neurologist 1 and 3 months after parturition. In addition, 20 randomly sampled participants who had denied any symptoms in the screening questions were interviewed as controls. Women who fulfilled the IRLSSG criteria and had symptoms for at least 4 days/month were diagnosed as having RLS. RLS was further divided into transient or chronic forms, which were defined by the
absence or presence of symptoms 3 months after delivery, respectively. SAS version 9.2 for Windows was used for statistical analysis. For continuous variables, the Student’s t test was used to compare the means. For categorical variables the v2 test was used to test the difference between groups. We used the Mann–Whitney U test or Fisher exact test when analyzing the smaller sample subgroup. Multivariate analysis was also performed using logistic regression models with RLS during pregnancy (yes/no) as the response variable and RLS predictors found significant by univariate analysis as covariates. Each covariate was tested independently and with the main interaction terms. Measures of association were odds ratios (ORs) with 95% confidence intervals (95% CIs). The continuous variables were expressed as the mean ± SD. All calculated p values were two-tailed, and we considered p < 0.05 to be statistically significant. 3. Results 3.1. Study participants During the study period, 660 pregnant women fulfilled the requirements of the inclusion and exclusion criteria. After obtaining informed consent and the completed questionnaires, 461 questionnaires were considered valid. The principal demographic and clinical characteristics of participants with and without RLS during pregnancy are listed in Table 1. 3.2. Comparisons of participants with and without RLS during pregnancy The frequency of RLS among pregnant participants was 10.4% (n = 48). Age, BMI, education level, race/ethnicity, gravida, and number of parities were not different between the RLS and nonRLS groups. Further, exercise habits, smoking, alcohol consumption, and coffee drinking history were not different between these two groups. Pregnant women without RLS had a higher overall rate of folic acid and iron supplement consumption (defined as self-report of at least twice-weekly supplementation for a month during the first and second trimester of pregnancy) than those with RLS. Participants with RLS reported more comorbidities, such as anemia, thyroid disease, or peptic ulcer disease, than those without RLS. Pregnant women without RLS had significantly higher serum hemoglobin levels than those with RLS (12.6 ± 1.74 mg/dL versus 11.6 ± 1.86 mg/dL, p = 0.001). Table 2 shows the results of multiple logistic regression analyses, listing significant covariates that were identified by univariate analysis. Multivariate analysis identified anemia (OR: 2.91, 95% CI: 1.47–5.76) and peptic ulcer disease (OR: 11.65, 95% CI: 2.84–47.83) was independent predictors of RLS during pregnancy. Multivariate analysis also revealed a trend towards the association of alcohol consumption and pregnancy-related RLS, but this did not reach statistical significance (OR: 2.98, 95% CI: 0.99–9.02). 3.3. Comparisons of participants with transient and chronic RLS The comparisons of demographic data and medical history between pregnant women with chronic RLS and those with transient RLS are shown in Table 3. Among the participants with RLS, 13/48 (27.1%) had chronic RLS and 35 (72.9%) had transient RLS. The demographic characteristics of participants between both groups were not significantly different. Further, exercise habits, smoking history, and alcohol consumption history did not differ between these two groups. Participants with transient RLS during pregnancy reported more coffee consumption before pregnancy than
P.-H. Chen et al. / Sleep Medicine 13 (2012) 1153–1157 Table 1 Demographic characteristics of study participants stratified according to the presence or absence of RLS during pregnancy.a Characteristic
b
Age (years) BMI <18.5 18.5 6 BMI < 24 24 6 BMI < 27 P27
pc
RLS during pregnancy No (n = 413)
Yes (n = 48)
31.9 ± 4.8
32.1 ± 4.4
b
0.774 0.963
43 (10.4) 270 (65.4) 61 (14.8) 39 (9.4)
5 (10.4) 33 (68.8) 6 (12.5) 4 (8.3)
Educational level (years) 612 151 (36.6) >12 262 (63.4)
12 (25.0) 36 (75.0)
Gravida 1 2 P3
23 (47.9) 9 (18.8) 16 (33.3)
0.113
1155
3.4. Comparisons of sleep quality between participants with chronic or transient RLS during pregnancy In RLS sufferers, 81.2% (39/48) had sleep disturbance, and all of the women with chronic RLS were poor sleepers. Affected women reported lower total sleep time, longer sleep latency, frequent insomnia, and overall poorer sleep quality, with the highest degree of severity in the third trimester. Table 4 demonstrated the results of sleep qualities stratified according to the diagnosis of chronic or transient RLS. Participants with chronic RLS had worse sleep latency, sleep duration, and sleep efficiency in comparison to those with transient RLS.
0.219 174 (42.1) 127 (30.8) 112 (27.1)
Parity 1 2 P3
220 (53.3) 147 (35.6) 46 (11.1)
29 (60.4) 12 (25.0) 7 (14.6)
Exercise No Yes
249 (60.3) 164 (39.7)
29 (60.4) 19 (39.6)
Smoking before pregnancy No 383 (92.7) Yes 30 (7.3)
44 (91.7) 4 (8.3)
Alcohol before pregnancy No 395 (95.6) Yes 18 (4.4)
43 (89.6) 5 (10.4)
Coffee before pregnancy No Yes
101 (24.5) 312 (75.5)
7 (14.6) 41 (85.4)
Folic acidd No Yes Unknown
96 (23.2) 306 (74.1) 11 (2.7)
17 (35.4) 27 (56.3) 4 (8.3)
Irond No Yes Unknown
108 (26.2) 275 (66.6) 30 (7.3)
23 (47.9) 22 (45.8) 3 (6.3)
Anemiad No Yes
352 (85.2) 61 (14.8)
31 (65.6) 17 (35.4)
Thyroid diseased No Yes
406 (98.3) 7 (1.7)
45 (93.8) 3 (6.3)
Hepatitis No Yes
400 (96.9) 13 (3.2)
47 (97.9) 1 (2.1)
Diabetes No Yes
408 (98.8) 5 (1.2)
47 (97.9) 1 (2.1)
Peptic ulcerd No Yes
408 (98.8) 5 (1.2)
44 (91.7) 4 (8.3)
4. Discussion 0.325
0.987
0.789
0.068
0.126
0.013
0.006
<0.001
0.040
0.684
0.614
0.001
Abbreviations: BMI, body mass index; RLS, restless legs syndrome. a Table values are represented as mean ± SD for continuous variables and n (%) for categorical variables. b Percentages may not sum to 100% due to rounding. c The indicated p-values are based on the t test (continuous variables) or the v2 test (categorical variables). d p < 0.05.
those with chronic RLS. Folic acid and iron supplement consumption and history of comorbidities also did not differ between these two groups.
We found an overall 10.4% prevalence of RLS among the pregnant women in our study, and 2.8% were categorized as having chronic RLS. This figure was near the lower end of previously reported ranges of RLS prevalence in pregnancy. Five of the 13 women with chronic RLS (38.5%) developed symptoms for the first time during the index pregnancy. Therefore, RLS symptoms experienced during pregnancy may persist after delivery due to the unmasking of chronic RLS. This is consistent with a previous report that identified RLS during pregnancy as a risk factor for the development of chronic RLS [22]. Only 2.4% (11/461) of the participants, and 10% (5/48) of the affected women, were previously aware of RLS. A previous study has also shown that RLS was frequently unrecognized and undiagnosed in pregnancy [23]. A lower tendency among Taiwanese pregnant women to report subjective symptoms and a huge gap in knowledge regarding RLS among physicians might have led to the lower estimation of RLS prevalence in Taiwan. The above reasons are also likely to have contributed to the lower prevalence rates of family history of RLS in our study. A previous study demonstrated that the risk of RLS increased gradually for women with the number of parities [24]. In our study, there is no significant difference in the numbers of previous pregnancies between participants with and without RLS. This, however, may also be a problem of sample size, but, clearly, any effect of an increased number of pregnancies is much smaller than that of the first pregnancy. Some studies have shown significantly lower hemoglobin values in women with pregnancy-related RLS [9,13]; we observed a similar association. We estimated that iron and folate deficiency were correlated with the incidence of RLS. In our study, the RLS group had a significantly higher rate of anemia, as well as a lower rate of iron and folate supplementation, than the group without RLS. Previous studies proposed a relationship between hormone changes and the pathophysiology of RLS. Due to ethical limits, there is a lack of blood sampling in our study; however, women with RLS in our study showed a higher prevalence of thyroid disorders than those without RLS. Gastrointestinal disorders such as previous gastric surgery, celiac disease, and irritable bowel syndrome have been suggested as secondary causes of RLS [25–27]. Our questionnaire included a simple question regarding the history of physician-confirmed peptic ulcer disease and found that peptic ulcer disease itself could predispose pregnant women to develop RLS. Participants with chronic RLS were slightly older than those with transient RLS and those without RLS (mean ages, 33.7, 31.5, and 31.9, respectively). Fifty-four percent (26/48) of the women with RLS began their symptoms during the third trimester; 85% (22/26) of them were diagnosed with transient RLS. Only 8% of the affected women began their symptoms during the first trimester, and they were all diagnosed as having chronic RLS. Symptoms of most of the participants with transient RLS resolved promptly
1156
P.-H. Chen et al. / Sleep Medicine 13 (2012) 1153–1157
Table 2 Bivariate and multivariate associations between study variables and the presence of RLS. Characteristic
N
% RLS
Unadjusted OR (95% CI)
Adjusted OR (95% CI)
Alcohol consumption before pregnancy No 438 9.8 1.00 Yes 23 21.7 2.55 (0.90–7.22)
1.00 2.98 (0.99–9.02)
Folic acid No Yes
113 333
15.0 8.1
1.00 0.45 (0.24–0.83)
1.00 0.84 (0.28–2.50)
Iron No Yes
131 297
17.6 7.4
1.00 0.43 (0.23–0.78)
1.00 0.50 (0.17–1.45)
Anemia No Yes
383 78
8.1 21.8
1.00 3.17 (1.65–6.07)
1.00 2.91 (1.47–5.76)
Thyroid disease No 451 Yes 10
10.0 30.0
1.00 3.87 (0.97–15.48)
1.00 3.89 (0.88–17.25)
Ulcer No Yes
9.7 44.4
1.00 7.42 (1.92–28.66)
1.00 11.65 (2.84–47.83)
452 9
Abbreviation: RLS, restless legs syndrome.
within a week after delivery; only one affected woman claimed her symptoms persisted for a month. Although caffeine is known to be associated with RLS [28], no study has yet reported the significance of caffeine in obstetric RLS. We found that the regular use of coffee was significantly correlated with transient RLS, but not with chronic RLS. This interesting and important finding is sufficient to support further study of the correlation between caffeine and RLS in pregnancy, but insufficient to reach a conclusion at present. Sleep disturbance is known to be a common symptom among pregnant women, especially in the third trimester. Various factors might contribute to sleep disturbance in pregnant women, such as nocturia, difficulty in assuming the habitual sleep posture because of abdominal enlargement, fetal movement, lower back pain, and hormonal effects. A previous study [29] revealed that 60% of second- and third-trimester pregnant women in Taiwan had sleep disturbance, defined by a PSQI score >5. In our study, 81.2% (39/48) of RLS sufferers reported sleep disturbance, and all of the women with chronic RLS were poor sleepers. Participants with chronic RLS reported an overall poorer sleep quality than those with transient RLS. Our study has some limitations. First, this is a single urban tertiary care center-based study in the private sector, and, therefore, it may not be truly representative of the population. However, in Taiwan, the government-run, single payer insurance system covers more than 96% of citizens. Pregnant women have the right to choose any hospital or physician they wish. More than two-thirds of pregnant women birth in a hospital, and home birth is extreme rare. Fewer than 3% of our participants are referred. Second, this study mostly relied on self-report for documentation of vitamin/ mineral supplement consumption and past medical diseases. Self-reported information may lead to misdiagnoses due to poor patient–clinician communication, or the ‘‘health literacy’’ of the patient. Fortunately, most of the study participants were actually healthy. Anemia, known to be common but frequently undiagnosed during pregnancy, was confirmed by electronic medical records. Third, we utilized the John Hopkins RLS Severity Rating Scale to assess the severity of RLS. This one-question scale is probably not very sensitive for treatment evaluation. However, it has been validated against objective measures of RLS severity and sleep efficiency [19]. Therefore, the scale is useful in screening
Table 3 Description of the sample according to whether subjects have chronic or transient RLS during pregnancy.a Characteristic
Transient (n = 35) Age (years) BMI <18.5 18.5 6 BMI < 24 24 6 BMI < 27 P27
pc
RLS during pregnancy
31.5 ± 4.9
b
Chronic (n = 13) 33.7 ± 2.3
b
0.086 0.677
4 (11.4) 24 (68.6) 5 (14.3) 2 (5.7)
1 9 1 9
(7.7) (69.2) (7.7) (15.4)
Educational level (years) 612 10 (28.6) >12 25 (71.4)
2 (15.4) 11 (84.6)
0.348
Gravida 1 2 P3
19 (54.3) 5 (14.3) 11 (31.4)
4 (30.8) 4 (30.8) 5 (38.5)
Para 1 2 P3
22 (62.9) 7 (20.0) 6 (17.1)
7 (53.9) 5 (38.5) 1 (7.7)
Exercise Yes No
13 (37.1) 22 (62.9)
6 (46.2) 7 (53.9)
Smoke_b No Yes
32 (91.4) 3 (8.6)
12 (92.3) 1 (7.7)
0.271
0.368
0.571
0.922
Alcohol consumption before pregnancy No 31 (88.6) Yes 4 (11.4) Coffee consumption before pregnancy No 2 (5.7) Yes 33 (94.3)
0.707 12 (92.3) 1 (7.7)
d
0.004 5 (38.5) 8 (61.5)
Folic acid No Yes Unknown
12 (34.3) 19 (54.3) 4 (11.4)
5 (38.5) 8 (61.5) 0 (0.0)
0.445
Iron No Yes Unknown
16 (45.7) 16 (45.7) 3 (8.6)
7 (53.9) 6 (46.2) 0 (0.0)
Anemia No Yes
22 (62.9) 13 (37.1)
9 (69.2) 4 (30.8)
Thyroid disease No Yes
34 (97.1) 1 (2.9)
11 (84.6) 2 (15.4)
Hepatitis No Yes
35 (100.0) 0 (0.0)
12 (92.3) 1 (7.7)
Diabetes No Yes
34 (97.1) 1 (2.9)
13 (100.0) 0 (0.0)
Peptic ulcer No Yes
33 (94.3) 2 (5.7)
11 (84.6) 2 (15.4)
0.537
0.682
0.111
0.097
0.538
0.281
Abbreviation: RLS, restless legs syndrome. a Table values are represented as mean ± SD for continuous variables and n (%) for categorical variables. b Percentages may not sum to 100% due to rounding. c The indicated p-values are based on the Mann–Whitney U test (continuous variables) or the Fisher exact test (categorical variables). d p < 0.05.
and epidemiological research [30]. Fourth, to confirm the diagnosis of RLS, participants were interviewed either by phone or face-toface. To enhance the accuracy of diagnosis, we utilized the vali-
P.-H. Chen et al. / Sleep Medicine 13 (2012) 1153–1157 Table 4 Pittsburgh Sleep Quality Index (PSQI) scores stratified according to the presence of chronic or transient RLS.a PSQI mean component score
RLS during pregnancy
Sleep quality (0–3) Sleep latency (0–3) c Sleep duration (0–3) c Sleep efficiency (0–3) c Sleep disturbance (0–3) Sleep medication (0–3) Daytime dysfunction (0– 3) Sum (0–21) c Poor sleep (PSQI > 5) (n, %)c
2.0 ± 0.7 1.8 ± 0.9 1.1 ± 1.1 1.0 ± 1.2 1.9 ± 0.7 0.1 ± 0.3 0.9 ± 0.9
Transient RLS (n = 35)
8.7 ± 3.8 26 (74.3%)
pb Chronic RLS (n = 13) 2.4 ± 0.7 2.5 ± 0.9 2.1 ± 1.0 2.2 ± 1.1 2.2 ± 0.7 0.0 ± 0.0 0.8 ± 0.7
0.120 0.035 0.007 0.003 0.173 0.542 0.813
12.2 ± 3.0 13 (100%)
0.005 0.043
Abbreviations: RLS, restless legs syndrome; PSQI, Pittsburgh Sleep Quality Index. a Table values are mean ± SD for continuous variables. b The indicated p-values are based on the Mann–Whitney U test (continuous variables) or the Fisher exact test (categorical variables). c p < 0.05.
dated Hopkins telephone diagnostic interview to exclude RLS mimics. Finally, we followed participants up to 3 months after the delivery; therefore, we are unable to comment upon the long-term natural course of obstetric RLS based on our study. 5. Conclusion Our study confirms the high prevalence but underdiagnosis of RLS among pregnant women in Taiwan. A history of anemia and peptic ulcer disease are independent predictors of RLS during pregnancy. On the other hand, folate and iron supplements are protectors against obstetric RLS. More than half of RLS sufferers began their symptoms during the third trimester, and in affected women whose symptoms begin in the first trimester, a diagnosis of chronic RLS should be considered. Obstetric RLS might unmask chronic RLS if symptoms were not resolved within a month after delivery. The regular use of coffee is correlated with transient RLS, but not with chronic RLS. Women with chronic RLS reported poorer sleep quality overall than those with transient RLS. Conflict of interest The ICMJE Uniform Disclosure Form for Potential Conflicts of Interest associated with this article can be viewed by clicking on the following link: http://dx.doi.org/10.1016/j.sleep.2012.06.008.
Acknowledgment We would like to thank the Obstetrics Department at Mackay Memorial Hospital for allowing us to use the facility for our research. References [1] Coccagna G, Vetrugno R, Lombardi C, Provini F. Restless legs syndrome: an historical note. Sleep Med 2004;5:279–83.
1157
[2] Allen RP, Picchietti D, Hening WA, Trenkwalder C, Walters AS, Montplaisi J, et al. Restless legs syndrome: diagnostic criteria, special considerations, and epidemiology: a report from the restless legs syndrome diagnosis and epidemiology workshop at the National Institutes of Health. Sleep Med 2003;4:101–19. [3] Yeh P, Walters AS, Tsuang JW. Restless legs syndrome: a comprehensive overview on its epidemiology, risk factors, and treatment. Sleep Breath 2011. Epub ahead of print. [4] Chen PH, Cheng SJ. Restless legs syndrome among the elderly. Int J Gerontol 2009;3:197–203. [5] Chen NH, Chuang LP, Yang CT, Kushida CA, Hsu SC, Wang PC, et al. The prevalence of restless legs syndrome in Taiwanese adults. Psychiatry Clin Neurosci 2010;64:170–8. [6] Harvey Simon. 2009. Retrieved December 20, 2011, from American Accreditation HealthCare Commission website:. [7] Ekbom KA. Restless legs syndrome. Acta Med Scand 1945;158(Suppl.):4–122. [8] Ismailogullari S, Ozturk A, Mazicioglu MM, Serin S, Gultekin M, Aksu M. Restless legs syndrome and pregnancy in Kayseri, Turkey: a hospital based survey. Sleep Biol Rhythms 2010;8:137–43. [9] Sikandar R, Khealani BA, Wasay M. Predictors of restless legs syndrome in pregnancy: a hospital based cross sectional survey from Pakistan. Sleep Med 2009;10:676–8. [10] Patrick LR. Restless legs syndrome: pathophysiology and the role of iron and folate. Altern Med Rev 2007;12:101–12. [11] Dzaja A, Wehrle R, Lancel M, Pollmächer T. Elevated estradiol plasma levels in women with restless legs during pregnancy. Sleep 2009;32:169–74. [12] Pereira Jr JC, Pradella-Hallinan M, Lins Pessoa H. Imbalance between thyroid hormones and the dopaminergic system might be central to the pathophysiology of restless legs syndrome: a hypothesis. Clinics 2010;65:548–54. [13] Tunç T, Karadag˘ YS, Dog˘ulu F, Inan LE. Predisposing factors of restless legs syndrome in pregnancy. Mov Disord 2007;22:627–31. [14] Lee KA, Zaffke ME, Baratte-Beebe K. Restless legs syndrome and sleep disturbance during pregnancy: the role of folate and iron. J Womens Health Gend Based Med 2001;10:335–41. [15] Manconi M, Govoni V, De Vito A, Economou NT, Cesnik E, Casetta I, et al. Restless legs syndrome and pregnancy. Neurology 2004;63(6):1065–9. [16] Neau JP, Marion P, Mathis S, Julian A, Godeneche G, Larrieu D, et al. Restless legs syndrome and pregnancy: follow-up of pregnant women before and after delivery. Eur Neurol 2010;64(6):361–6. [17] Uglane MT, Westad S, Backe B. Restless legs syndrome in pregnancy is a frequent disorder with a good prognosis. Acta Obstet Gynecol Scand 2011;90(9):1046–8. [18] Hening WA, Allen RP, Washburn M, Lesage S, Earley CJ. Validation of the Hopkins telephone diagnostic interview for restless legs syndrome. Sleep Med 2008;9:283–9. [19] Allen RP, Earley CJ. Validation of the John Hopkins restless legs severity scale. Sleep Med 2001;2:239–42. [20] Tsai PS, Wang SY, Wang MY, Su CT, Yang TT, Huang CJ, et al. Psychometric evaluation of the Chinese version of the Pittsburgh Sleep Quality Index (CPSQI) in primary insomnia and control subjects. Qual Life Res 2005;14:1943–52. [21] Centers for Disease Control (CDC). CDC criteria for anemia in children and childbearing-aged women. MMWR Morb Mortal Wkly Rep 1989;38:400–4. [22] Cesnik E, Casetta I, Turri M, Govoni V, Granieri E, Strambi LF, et al. Transient RLS during pregnancy is a risk factor for the chronic idiopathic form. Neurology 2010;75:2117–20. [23] Nasir A, Khuwaja AK. Restless legs syndrome: common but frequently unrecognized disorder in pregnancy. J Pak Med Assoc 2007;57:164. [24] Berger K, Luedemann J, Trenkwalder C, John U, Kessler C. Sex and the risk of restless legs syndrome in the general population. Arch Intern Med 2004;164:196–202. [25] Banerji NK, Hurwitz LJ. Restless legs syndrome, with particular reference to its occurrence after gastric surgery. Br Med J 1970;4:774–5. [26] Moccia M, Pellecchia MT, Erro R, Zingone F, Marelli S, Barone DG, et al. Restless legs syndrome is a common feature of adult celiac disease. Mov Disord 2010;25:877–81. [27] Weinstock LB, Walters AS. Restless legs syndrome is associated with irritable bowel syndrome and small intestinal bacterial overgrowth. Sleep Med 2011;12:610–3. [28] Lutz EG. Restless legs, anxiety and caffeinism. J Clin Psychiatry 1978;39:693–8. [29] Ko SH, Chang SC, Chen CH. A comparative study of sleep quality between pregnant and nonpregnant Taiwanese women. J Nurs Scholarsh 2010;42:23–30. [30] Kohnen R, Allen RP, Benes H, Garcia-Borreguero D, Hening WA, Stiasny-Kolster K, et al. Assessment of restless legs syndrome–methodological approaches for use in practice and clinical trials. Mov Disord 2007;22:S485–94.
Contents lists available at SciVerse ScienceDirect
Sleep Medicine journal homepage: www.elsevier.com/locate/sleep
Original Article
Risk factors and prevalence rate of restless legs syndrome among pregnant women in Taiwan Pei-Hao Chen a,c,d, Kuang-Chung Liou a, Chie-Pein Chen b, Shih-Jung Cheng a,d,⇑ a
Department of Neurology, Mackay Memorial Hospital, Taipei, Taiwan Division of High Risk Pregnancy, Mackay Memorial Hospital, Taipei, Taiwan c Graduate Institute of Mechanical and Electrical Engineering, National Taipei University of Technology, Taipei, Taiwan d Mackay Medicine, Nursing and Management College, Taipei, Taiwan b
a r t i c l e
i n f o
Article history: Received 10 February 2012 Received in revised form 8 June 2012 Accepted 15 June 2012 Available online 31 July 2012 Keywords: Anemia Coffee Peptic ulcer disease Pregnancy Restless legs syndrome Sleep disorders
a b s t r a c t Background: The goal of this study was to assess the prevalence and clinical correlates of restless legs syndrome (RLS) among pregnant Taiwanese women. Methods: We enrolled 461 pregnant women (18–45 years) admitted at Mackay Memorial Hospital for delivery from September 2010 to May 2011. The face-to-face questionnaire used to gather data included assessment of RLS diagnostic criteria, and questions related to RLS. Results: The overall prevalence rate of RLS among the study participants was 10.4%; 2.8% were categorized as having chronic RLS. Participants without RLS reported higher folate and iron supplement consumption than those with RLS. Multivariate analysis revealed significant associations of RLS with anemia and peptic ulcer disease. Participants with transient RLS during pregnancy reported more regular coffee consumption before pregnancy, and better sleep latency, duration, and efficiency, than those with chronic RLS. Overall, 81.2% of RLS sufferers reported sleep disturbances. Conclusions: Our study revealed highly prevalent but poorly recognized RLS among Taiwanese pregnant women. The identification of predictors such as medical comorbidities, and protectors such as folate and iron supplements, is warranted for obstetric RLS. In most cases, symptoms began during the second or third trimester and resolved within a week after delivery. Restricted coffee consumption before pregnancy is encouraged, but further evidence is needed to support this recommendation. Ó 2012 Elsevier B.V. All rights reserved.
1. Introduction Restless legs syndrome (RLS), first given its name in 1945 by the Swedish neurologist Karl-Axel Ekbom [1], is a sleep and movement disorder characterized by unpleasant sensations in the legs occurring at rest and at night, with an irresistible urge to move. The diagnosis of RLS was upheld by the International Restless Legs Syndrome Study Group (IRLSSG), with four cardinal features based on the history of the patient [2]. It is one of the leading known causes of insomnia and common movement disorders. RLS affects 5–15% of the general population and it has an increased prevalence in older individuals [3,4]. The prevalence of RLS in Taiwanese adults was found to be 1.57%, suggesting substantially lower rates in Asian populations compared with Caucasians [5]. Collectively, findings to date offer evidence for an association between RLS and medical comorbidities such as osteoarthritis, varicose veins, type 2 diabe⇑ Corresponding author. Address: Department of Neurology, Mackay Memorial Hospital, No. 92, Sec. 2, Zhongshan N. Rd., Zhongshan Dist., Taipei City 10449, Taiwan. Tel.: +886 225433535x2316. E-mail address: [email protected] (S.-J. Cheng). 1389-9457/$ - see front matter Ó 2012 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.sleep.2012.06.008
tes, hypertension, hypothyroidism, fibromyalgia, rheumatoid arthritis, emphysema, chronic alcoholism, sleep apnea, chronic headaches, and injuries or diseases that affect the central nervous system [6]. However, the literature in this area is sparse and less consistent, so the possible relationships between these factors and RLS are not clearly defined. The development of RLS during pregnancy has been long known; Ekbom noted in his original report that 11.3% of 486 pregnant women had RLS [7]. Thus, the prevalence of RLS among pregnant women was twice that reported by Ekbom among a sample of 230 healthy women (5.7%). Studies found that RLS prevalence rates in pregnant women are similar in Western and Eastern countries [3]. To the best of our knowledge, overall reported prevalence rates of RLS in pregnant women based on IRLSSG criteria with face-toface interviews range from 10.5% in Turkey to 30% in Pakistan [8,9]. The main causes of association between RLS and pregnancy are not well known. Hypotheses for the higher prevalence of RLS during pregnancy involve reduced serum iron or folate levels and hormonal changes such as estrogen, progesterone, prolactin, and thyroid hormones [10–13]. However, previous studies have shown
1154
P.-H. Chen et al. / Sleep Medicine 13 (2012) 1153–1157
inconsistent results with respect to their close relationships [13– 17]. To date, there have been no reports on the prevalence of RLS among Taiwanese pregnant women. In this study, the purposes are 2-fold: (1) to determine the prevalence of RLS in Taiwanese pregnant women, and (2) to explore the clinical correlates of RLS in pregnancy and its impact on sleep.
2. Methods We prospectively enrolled pregnant women aged P18 years who were admitted to the Department of Obstetrics of Mackay Memorial Hospital for delivery from September 2010 to May 2011. Women were not included in the study if they had eclampsia, preeclampsia, or severe ankle edema. Specific diseases known to be causes of symptomatic RLS were not excluded. This study was approved by the Mackay Memorial Hospital Institutional Review Board and informed consent was obtained from each participant. A questionnaire was administered during the face-to-face interview. This questionnaire consisted of 20 questions, including demographic characteristics (age, height, body weight, education, race, number of previous pregnancies, number of children, etc.), personal behavior (exercise, smoking, drinking, caffeine, iron and vitamin supplement consumption), and past medical illnesses. Medical information was also obtained from the medical records of all participants, including demographic characteristics, previous pregnancies, drug history, and sleep disorders. Diagnosis of RLS was established on the basis of all four of the IRLSSG criteria. All participants were evaluated by the same neurologist (Chen P.H.), who reviewed all the completed questionnaires and interviewed the women either by phone or face-to-face if they responded positively to at least one of the four screening questions based on the IRLSSG criteria. There is a possibility of confusion between the symptoms of RLS and those of positional discomfort, hypnagogic jerks, nocturnal leg cramps, or musculoskeletal disorders. To prevent this confusion, we utilized the validated Hopkins telephone diagnostic interview [18] that contained 11 specific questions to confirm RLS and to eliminate potential false positives. Participants who fulfilled the IRLSSG criteria were asked to fill in a form that included questions to assess their RLS history and the RLS history of their first-degree relatives, as well as questions based on the John Hopkins RLS Severity Rating Scale [19] and the Chinese version of Pittsburgh Sleep Quality Index (PSQI) [20]. The John Hopkins RLS Severity Rating Scale consists of one question regarding the usual (>50% of days) time of the day that symptoms started. Severity of RLS can be classified as minimal (score 0), mild (score 1), moderate (score 2), and severe (score 3). This crude severity rating scale rates most people with RLS fairly well, but it does not take into account the intensity of the RLS symptoms and their effect on daytime functioning. The Pittsburgh Sleep Quality Index measures components of sleep quality, sleep latency, sleep duration, habitual sleep efficiency, sleep disturbances, use of sleeping medications, daytime dysfunction, and a global score. Scores range from 0 to 21, with scores greater than five indicating clinically significant sleep disturbance. Hemoglobin levels were obtained from routine blood tests between week 24 and week 28 of pregnancy. Women were considered anemic if hemoglobin levels were below 11 g/dL [21]. All participants affected by RLS underwent follow-up telephone interviews conducted by the same neurologist 1 and 3 months after parturition. In addition, 20 randomly sampled participants who had denied any symptoms in the screening questions were interviewed as controls. Women who fulfilled the IRLSSG criteria and had symptoms for at least 4 days/month were diagnosed as having RLS. RLS was further divided into transient or chronic forms, which were defined by the
absence or presence of symptoms 3 months after delivery, respectively. SAS version 9.2 for Windows was used for statistical analysis. For continuous variables, the Student’s t test was used to compare the means. For categorical variables the v2 test was used to test the difference between groups. We used the Mann–Whitney U test or Fisher exact test when analyzing the smaller sample subgroup. Multivariate analysis was also performed using logistic regression models with RLS during pregnancy (yes/no) as the response variable and RLS predictors found significant by univariate analysis as covariates. Each covariate was tested independently and with the main interaction terms. Measures of association were odds ratios (ORs) with 95% confidence intervals (95% CIs). The continuous variables were expressed as the mean ± SD. All calculated p values were two-tailed, and we considered p < 0.05 to be statistically significant. 3. Results 3.1. Study participants During the study period, 660 pregnant women fulfilled the requirements of the inclusion and exclusion criteria. After obtaining informed consent and the completed questionnaires, 461 questionnaires were considered valid. The principal demographic and clinical characteristics of participants with and without RLS during pregnancy are listed in Table 1. 3.2. Comparisons of participants with and without RLS during pregnancy The frequency of RLS among pregnant participants was 10.4% (n = 48). Age, BMI, education level, race/ethnicity, gravida, and number of parities were not different between the RLS and nonRLS groups. Further, exercise habits, smoking, alcohol consumption, and coffee drinking history were not different between these two groups. Pregnant women without RLS had a higher overall rate of folic acid and iron supplement consumption (defined as self-report of at least twice-weekly supplementation for a month during the first and second trimester of pregnancy) than those with RLS. Participants with RLS reported more comorbidities, such as anemia, thyroid disease, or peptic ulcer disease, than those without RLS. Pregnant women without RLS had significantly higher serum hemoglobin levels than those with RLS (12.6 ± 1.74 mg/dL versus 11.6 ± 1.86 mg/dL, p = 0.001). Table 2 shows the results of multiple logistic regression analyses, listing significant covariates that were identified by univariate analysis. Multivariate analysis identified anemia (OR: 2.91, 95% CI: 1.47–5.76) and peptic ulcer disease (OR: 11.65, 95% CI: 2.84–47.83) was independent predictors of RLS during pregnancy. Multivariate analysis also revealed a trend towards the association of alcohol consumption and pregnancy-related RLS, but this did not reach statistical significance (OR: 2.98, 95% CI: 0.99–9.02). 3.3. Comparisons of participants with transient and chronic RLS The comparisons of demographic data and medical history between pregnant women with chronic RLS and those with transient RLS are shown in Table 3. Among the participants with RLS, 13/48 (27.1%) had chronic RLS and 35 (72.9%) had transient RLS. The demographic characteristics of participants between both groups were not significantly different. Further, exercise habits, smoking history, and alcohol consumption history did not differ between these two groups. Participants with transient RLS during pregnancy reported more coffee consumption before pregnancy than
P.-H. Chen et al. / Sleep Medicine 13 (2012) 1153–1157 Table 1 Demographic characteristics of study participants stratified according to the presence or absence of RLS during pregnancy.a Characteristic
b
Age (years) BMI <18.5 18.5 6 BMI < 24 24 6 BMI < 27 P27
pc
RLS during pregnancy No (n = 413)
Yes (n = 48)
31.9 ± 4.8
32.1 ± 4.4
b
0.774 0.963
43 (10.4) 270 (65.4) 61 (14.8) 39 (9.4)
5 (10.4) 33 (68.8) 6 (12.5) 4 (8.3)
Educational level (years) 612 151 (36.6) >12 262 (63.4)
12 (25.0) 36 (75.0)
Gravida 1 2 P3
23 (47.9) 9 (18.8) 16 (33.3)
0.113
1155
3.4. Comparisons of sleep quality between participants with chronic or transient RLS during pregnancy In RLS sufferers, 81.2% (39/48) had sleep disturbance, and all of the women with chronic RLS were poor sleepers. Affected women reported lower total sleep time, longer sleep latency, frequent insomnia, and overall poorer sleep quality, with the highest degree of severity in the third trimester. Table 4 demonstrated the results of sleep qualities stratified according to the diagnosis of chronic or transient RLS. Participants with chronic RLS had worse sleep latency, sleep duration, and sleep efficiency in comparison to those with transient RLS.
0.219 174 (42.1) 127 (30.8) 112 (27.1)
Parity 1 2 P3
220 (53.3) 147 (35.6) 46 (11.1)
29 (60.4) 12 (25.0) 7 (14.6)
Exercise No Yes
249 (60.3) 164 (39.7)
29 (60.4) 19 (39.6)
Smoking before pregnancy No 383 (92.7) Yes 30 (7.3)
44 (91.7) 4 (8.3)
Alcohol before pregnancy No 395 (95.6) Yes 18 (4.4)
43 (89.6) 5 (10.4)
Coffee before pregnancy No Yes
101 (24.5) 312 (75.5)
7 (14.6) 41 (85.4)
Folic acidd No Yes Unknown
96 (23.2) 306 (74.1) 11 (2.7)
17 (35.4) 27 (56.3) 4 (8.3)
Irond No Yes Unknown
108 (26.2) 275 (66.6) 30 (7.3)
23 (47.9) 22 (45.8) 3 (6.3)
Anemiad No Yes
352 (85.2) 61 (14.8)
31 (65.6) 17 (35.4)
Thyroid diseased No Yes
406 (98.3) 7 (1.7)
45 (93.8) 3 (6.3)
Hepatitis No Yes
400 (96.9) 13 (3.2)
47 (97.9) 1 (2.1)
Diabetes No Yes
408 (98.8) 5 (1.2)
47 (97.9) 1 (2.1)
Peptic ulcerd No Yes
408 (98.8) 5 (1.2)
44 (91.7) 4 (8.3)
4. Discussion 0.325
0.987
0.789
0.068
0.126
0.013
0.006
<0.001
0.040
0.684
0.614
0.001
Abbreviations: BMI, body mass index; RLS, restless legs syndrome. a Table values are represented as mean ± SD for continuous variables and n (%) for categorical variables. b Percentages may not sum to 100% due to rounding. c The indicated p-values are based on the t test (continuous variables) or the v2 test (categorical variables). d p < 0.05.
those with chronic RLS. Folic acid and iron supplement consumption and history of comorbidities also did not differ between these two groups.
We found an overall 10.4% prevalence of RLS among the pregnant women in our study, and 2.8% were categorized as having chronic RLS. This figure was near the lower end of previously reported ranges of RLS prevalence in pregnancy. Five of the 13 women with chronic RLS (38.5%) developed symptoms for the first time during the index pregnancy. Therefore, RLS symptoms experienced during pregnancy may persist after delivery due to the unmasking of chronic RLS. This is consistent with a previous report that identified RLS during pregnancy as a risk factor for the development of chronic RLS [22]. Only 2.4% (11/461) of the participants, and 10% (5/48) of the affected women, were previously aware of RLS. A previous study has also shown that RLS was frequently unrecognized and undiagnosed in pregnancy [23]. A lower tendency among Taiwanese pregnant women to report subjective symptoms and a huge gap in knowledge regarding RLS among physicians might have led to the lower estimation of RLS prevalence in Taiwan. The above reasons are also likely to have contributed to the lower prevalence rates of family history of RLS in our study. A previous study demonstrated that the risk of RLS increased gradually for women with the number of parities [24]. In our study, there is no significant difference in the numbers of previous pregnancies between participants with and without RLS. This, however, may also be a problem of sample size, but, clearly, any effect of an increased number of pregnancies is much smaller than that of the first pregnancy. Some studies have shown significantly lower hemoglobin values in women with pregnancy-related RLS [9,13]; we observed a similar association. We estimated that iron and folate deficiency were correlated with the incidence of RLS. In our study, the RLS group had a significantly higher rate of anemia, as well as a lower rate of iron and folate supplementation, than the group without RLS. Previous studies proposed a relationship between hormone changes and the pathophysiology of RLS. Due to ethical limits, there is a lack of blood sampling in our study; however, women with RLS in our study showed a higher prevalence of thyroid disorders than those without RLS. Gastrointestinal disorders such as previous gastric surgery, celiac disease, and irritable bowel syndrome have been suggested as secondary causes of RLS [25–27]. Our questionnaire included a simple question regarding the history of physician-confirmed peptic ulcer disease and found that peptic ulcer disease itself could predispose pregnant women to develop RLS. Participants with chronic RLS were slightly older than those with transient RLS and those without RLS (mean ages, 33.7, 31.5, and 31.9, respectively). Fifty-four percent (26/48) of the women with RLS began their symptoms during the third trimester; 85% (22/26) of them were diagnosed with transient RLS. Only 8% of the affected women began their symptoms during the first trimester, and they were all diagnosed as having chronic RLS. Symptoms of most of the participants with transient RLS resolved promptly
1156
P.-H. Chen et al. / Sleep Medicine 13 (2012) 1153–1157
Table 2 Bivariate and multivariate associations between study variables and the presence of RLS. Characteristic
N
% RLS
Unadjusted OR (95% CI)
Adjusted OR (95% CI)
Alcohol consumption before pregnancy No 438 9.8 1.00 Yes 23 21.7 2.55 (0.90–7.22)
1.00 2.98 (0.99–9.02)
Folic acid No Yes
113 333
15.0 8.1
1.00 0.45 (0.24–0.83)
1.00 0.84 (0.28–2.50)
Iron No Yes
131 297
17.6 7.4
1.00 0.43 (0.23–0.78)
1.00 0.50 (0.17–1.45)
Anemia No Yes
383 78
8.1 21.8
1.00 3.17 (1.65–6.07)
1.00 2.91 (1.47–5.76)
Thyroid disease No 451 Yes 10
10.0 30.0
1.00 3.87 (0.97–15.48)
1.00 3.89 (0.88–17.25)
Ulcer No Yes
9.7 44.4
1.00 7.42 (1.92–28.66)
1.00 11.65 (2.84–47.83)
452 9
Abbreviation: RLS, restless legs syndrome.
within a week after delivery; only one affected woman claimed her symptoms persisted for a month. Although caffeine is known to be associated with RLS [28], no study has yet reported the significance of caffeine in obstetric RLS. We found that the regular use of coffee was significantly correlated with transient RLS, but not with chronic RLS. This interesting and important finding is sufficient to support further study of the correlation between caffeine and RLS in pregnancy, but insufficient to reach a conclusion at present. Sleep disturbance is known to be a common symptom among pregnant women, especially in the third trimester. Various factors might contribute to sleep disturbance in pregnant women, such as nocturia, difficulty in assuming the habitual sleep posture because of abdominal enlargement, fetal movement, lower back pain, and hormonal effects. A previous study [29] revealed that 60% of second- and third-trimester pregnant women in Taiwan had sleep disturbance, defined by a PSQI score >5. In our study, 81.2% (39/48) of RLS sufferers reported sleep disturbance, and all of the women with chronic RLS were poor sleepers. Participants with chronic RLS reported an overall poorer sleep quality than those with transient RLS. Our study has some limitations. First, this is a single urban tertiary care center-based study in the private sector, and, therefore, it may not be truly representative of the population. However, in Taiwan, the government-run, single payer insurance system covers more than 96% of citizens. Pregnant women have the right to choose any hospital or physician they wish. More than two-thirds of pregnant women birth in a hospital, and home birth is extreme rare. Fewer than 3% of our participants are referred. Second, this study mostly relied on self-report for documentation of vitamin/ mineral supplement consumption and past medical diseases. Self-reported information may lead to misdiagnoses due to poor patient–clinician communication, or the ‘‘health literacy’’ of the patient. Fortunately, most of the study participants were actually healthy. Anemia, known to be common but frequently undiagnosed during pregnancy, was confirmed by electronic medical records. Third, we utilized the John Hopkins RLS Severity Rating Scale to assess the severity of RLS. This one-question scale is probably not very sensitive for treatment evaluation. However, it has been validated against objective measures of RLS severity and sleep efficiency [19]. Therefore, the scale is useful in screening
Table 3 Description of the sample according to whether subjects have chronic or transient RLS during pregnancy.a Characteristic
Transient (n = 35) Age (years) BMI <18.5 18.5 6 BMI < 24 24 6 BMI < 27 P27
pc
RLS during pregnancy
31.5 ± 4.9
b
Chronic (n = 13) 33.7 ± 2.3
b
0.086 0.677
4 (11.4) 24 (68.6) 5 (14.3) 2 (5.7)
1 9 1 9
(7.7) (69.2) (7.7) (15.4)
Educational level (years) 612 10 (28.6) >12 25 (71.4)
2 (15.4) 11 (84.6)
0.348
Gravida 1 2 P3
19 (54.3) 5 (14.3) 11 (31.4)
4 (30.8) 4 (30.8) 5 (38.5)
Para 1 2 P3
22 (62.9) 7 (20.0) 6 (17.1)
7 (53.9) 5 (38.5) 1 (7.7)
Exercise Yes No
13 (37.1) 22 (62.9)
6 (46.2) 7 (53.9)
Smoke_b No Yes
32 (91.4) 3 (8.6)
12 (92.3) 1 (7.7)
0.271
0.368
0.571
0.922
Alcohol consumption before pregnancy No 31 (88.6) Yes 4 (11.4) Coffee consumption before pregnancy No 2 (5.7) Yes 33 (94.3)
0.707 12 (92.3) 1 (7.7)
d
0.004 5 (38.5) 8 (61.5)
Folic acid No Yes Unknown
12 (34.3) 19 (54.3) 4 (11.4)
5 (38.5) 8 (61.5) 0 (0.0)
0.445
Iron No Yes Unknown
16 (45.7) 16 (45.7) 3 (8.6)
7 (53.9) 6 (46.2) 0 (0.0)
Anemia No Yes
22 (62.9) 13 (37.1)
9 (69.2) 4 (30.8)
Thyroid disease No Yes
34 (97.1) 1 (2.9)
11 (84.6) 2 (15.4)
Hepatitis No Yes
35 (100.0) 0 (0.0)
12 (92.3) 1 (7.7)
Diabetes No Yes
34 (97.1) 1 (2.9)
13 (100.0) 0 (0.0)
Peptic ulcer No Yes
33 (94.3) 2 (5.7)
11 (84.6) 2 (15.4)
0.537
0.682
0.111
0.097
0.538
0.281
Abbreviation: RLS, restless legs syndrome. a Table values are represented as mean ± SD for continuous variables and n (%) for categorical variables. b Percentages may not sum to 100% due to rounding. c The indicated p-values are based on the Mann–Whitney U test (continuous variables) or the Fisher exact test (categorical variables). d p < 0.05.
and epidemiological research [30]. Fourth, to confirm the diagnosis of RLS, participants were interviewed either by phone or face-toface. To enhance the accuracy of diagnosis, we utilized the vali-
P.-H. Chen et al. / Sleep Medicine 13 (2012) 1153–1157 Table 4 Pittsburgh Sleep Quality Index (PSQI) scores stratified according to the presence of chronic or transient RLS.a PSQI mean component score
RLS during pregnancy
Sleep quality (0–3) Sleep latency (0–3) c Sleep duration (0–3) c Sleep efficiency (0–3) c Sleep disturbance (0–3) Sleep medication (0–3) Daytime dysfunction (0– 3) Sum (0–21) c Poor sleep (PSQI > 5) (n, %)c
2.0 ± 0.7 1.8 ± 0.9 1.1 ± 1.1 1.0 ± 1.2 1.9 ± 0.7 0.1 ± 0.3 0.9 ± 0.9
Transient RLS (n = 35)
8.7 ± 3.8 26 (74.3%)
pb Chronic RLS (n = 13) 2.4 ± 0.7 2.5 ± 0.9 2.1 ± 1.0 2.2 ± 1.1 2.2 ± 0.7 0.0 ± 0.0 0.8 ± 0.7
0.120 0.035 0.007 0.003 0.173 0.542 0.813
12.2 ± 3.0 13 (100%)
0.005 0.043
Abbreviations: RLS, restless legs syndrome; PSQI, Pittsburgh Sleep Quality Index. a Table values are mean ± SD for continuous variables. b The indicated p-values are based on the Mann–Whitney U test (continuous variables) or the Fisher exact test (categorical variables). c p < 0.05.
dated Hopkins telephone diagnostic interview to exclude RLS mimics. Finally, we followed participants up to 3 months after the delivery; therefore, we are unable to comment upon the long-term natural course of obstetric RLS based on our study. 5. Conclusion Our study confirms the high prevalence but underdiagnosis of RLS among pregnant women in Taiwan. A history of anemia and peptic ulcer disease are independent predictors of RLS during pregnancy. On the other hand, folate and iron supplements are protectors against obstetric RLS. More than half of RLS sufferers began their symptoms during the third trimester, and in affected women whose symptoms begin in the first trimester, a diagnosis of chronic RLS should be considered. Obstetric RLS might unmask chronic RLS if symptoms were not resolved within a month after delivery. The regular use of coffee is correlated with transient RLS, but not with chronic RLS. Women with chronic RLS reported poorer sleep quality overall than those with transient RLS. Conflict of interest The ICMJE Uniform Disclosure Form for Potential Conflicts of Interest associated with this article can be viewed by clicking on the following link: http://dx.doi.org/10.1016/j.sleep.2012.06.008.
Acknowledgment We would like to thank the Obstetrics Department at Mackay Memorial Hospital for allowing us to use the facility for our research. References [1] Coccagna G, Vetrugno R, Lombardi C, Provini F. Restless legs syndrome: an historical note. Sleep Med 2004;5:279–83.
1157
[2] Allen RP, Picchietti D, Hening WA, Trenkwalder C, Walters AS, Montplaisi J, et al. Restless legs syndrome: diagnostic criteria, special considerations, and epidemiology: a report from the restless legs syndrome diagnosis and epidemiology workshop at the National Institutes of Health. Sleep Med 2003;4:101–19. [3] Yeh P, Walters AS, Tsuang JW. Restless legs syndrome: a comprehensive overview on its epidemiology, risk factors, and treatment. Sleep Breath 2011. Epub ahead of print. [4] Chen PH, Cheng SJ. Restless legs syndrome among the elderly. Int J Gerontol 2009;3:197–203. [5] Chen NH, Chuang LP, Yang CT, Kushida CA, Hsu SC, Wang PC, et al. The prevalence of restless legs syndrome in Taiwanese adults. Psychiatry Clin Neurosci 2010;64:170–8. [6] Harvey Simon. 2009. Retrieved December 20, 2011, from American Accreditation HealthCare Commission website: