half of the night and then decreased to a level similar to the beginning of the ..... the half-life of the medication with the TON pattern may improve their efficacy.
Sleep. 15(4):306-311
© 1992 American Sleep Disorders Association and Sleep Research Society
Time-of-Night Patterns in PLMS Activity *William J. Culpepper, *Pietro Badia and tJoseph 1. Shaffer *Department of Psychology, Bowling Green State University, Bowling Green, Ohio and tUniversity of Toledo and St. Vincent Medical Center, Toledo, Ohio U.S.A.
Summary: Time-of-night patterns for periodic leg movement in sleep (PLMS) activity were assessed in 72 patients diagnosed as having PLMS (46 with only PLMS, 10 with narcolepsy and PLMS and 16 with sleep apnea and PLMS). Two distinct patterns ofPLMS activity were identified. For one pattern, PLMS activity was high initially upon falling asleep and then decreased across the remainder of the night with the majority of the activity occurring during the first half of the night. For the second pattern, PLMS activity was relatively evenly distributed across the night with slightly greater activity in the middle of the night. These patterns were highly reliable across two nights of recording for group 1 and less so for group 2 even though there was significant variability between nights in the number of PLMS. Patients in group 2 were younger (except for those with sleep apnea) and were sleepier during the day. Implications of these findings for pharmacologic management are discussed. Key Words: Time-of-nightPeriodic leg movements in sleep-Narcolepsy-Apne:a.
Periodic leg movements in sleep (PLMS) is a disorder characterized by rhythmical movements (jerks) of the lower legs recurring at 20-40-second intervals that may occur hundreds of times each night and are often associated with arousals and awakenings (1--4). This disorder may result in either a complaint of insomnia or excessive daytime sleepiness and is also found in asymptomatic individuals (i.e. no sleep/wake complaint) (1,5). The present report identifies patients with different patterns of PLMS activity across the night. A significant portion of the population appears to suffer from PLMS with the incidence of PLMS increasing with age (3,5-7) and in association with other sleep disorders (5,8) and with significant variability across nights in the number of leg movements (9). Other than the latter findings, little definitive data regarding PLMS exist and the etiology of this disorder remains unknown. As a consequence, treatment of PLMS has been primarily pharmacological and has met with limited success (10-12). Time-of-night (TON) changes in PLMS activity have been noted in the literature (13) with the conclusion that there is no relationship between PLMS and TON. As a consequence, TON patterns in PLMS activity Accepted for publication January 1992. Address correspondence and reprint requests to Pietro Badia, Department of Psychology, Bowling Green State University, Bowling Green, OR 43403, U.S.A.
were ignored until recently (14) when a difference in the pattern between insomnia and narcolepsy patients was reported: PLMS decreased across the night for the insomnia patients, whereas patients with narcolepsy showed a more stable pattern across the night. Independent of the latter study, we (15) reported findings on patients diagnosed with PLMS of two distinct patterns in PLMS activity as a function of TON: one group of patients showed a marked decrease in PLMS activity across the night, and another group of patients showed a small increase across the night. The present study is based on a larger sample that permitted assessment of TON patterns in patients with PLMS and in patients with PLMS and sleep apnea or narcolepsy. The objectives of this study were to assess 1) the reliability of TON patterns in PLMS activity across consecutive nights, 2) the differential effects ofthe TON pattern on sleep and sleepiness and 3) whether demographic variables including age, sex and waking complaint might distinguish between the two TON patterns. METHODS Subjects
Seventy-two patients evaluated at the St. Vincent Medical Center Sleep Disorders Center diagnosed with PLMS (PLMS index ::::: 5) without medical conditions known to exacerbate PLMS (e.g. diabetes, renal dis-
306
PATTERNS IN PLMS ACTIVITY
.'
"
"
ease, medications) were retrospectively evaluated. Of these 72 patients, 46 were diagnosed as having only PLMS (30 males, 16 females), 16 were diagnosed as having sleep apnea and PLMS (12 males, 4 females) and 10 were diagnosed as having narcolepsy and PLMS (5 males, 5 females). Each patient had two consecutive nights of polysomnography (PSG) and two series of multiple sleep latency tests (MSLTs). Standard procedures and recordings were employed that included the following measures: 1) electroencephalograph (EEG)-C3IC4 and 01/02, 2) left and right electrooculograph, 3) submental electromyograph (EMG), 4) electrocardiograph, 5) left and right anterior tibialis EMG, 6) thoracic effort and intercostal EMG, 7) oral/nasal airflow and 8) ear oximetery. Sleep staging and respiratory measures were scored according to standard procedures (16,17). The MSLT consisted of four naps at 1000, 1200, 1400 and 1600 hours (18). Each sleep record was scored in hourly blocks for the following dependent variables: 1) number of leg movements, 2) number of EMG (chin) arousals, 3) number ofEEG arousals (any change in the EMG/EEG at and up to 3 seconds following a leg movement), 4) number of stage shifts (including awakenings) and 5) the intermovement interval (IMI: mean time in seconds between movements within an episode). These measures were analyzed using a four-way group x diagnosis x night x TON mixed-factor ANOV A. Between-groups factors included group (1 and 2) and diagnosis (PLMS only, PLMS and narcolepsy, PLMS and apnea) whereas within (repeated)-group factors included night (1 and 2) and TON (1 through 8). Greenhouse-Geisser correction was used when the sphericity assumption was not met.
307
pattern of PLMS activity is remarkably stable from night 1 (N 1) to night 2 (N2). A significant group x TON interaction [F(7,924) > 6.30, p < 0.01] was found for PLMS, EMG arousals, EEG arousals and stage shifts. The TON main effect was significant for PLMS, EMG arousals and EEG arousals [F(7,924) > 5.06, p < 0.01], and the group main effect was significant for PLMS [F(1,132) = 6.56, p < 0.05]; group 1 patients displayed more PLMS than group 2 (317.6 ± 22.7 vs. 220.7 ± 26.0, respectively). No differences were found between nights for any measure of PLMS activity. IMI
Given the two PLMS patterns, we then assessed other measures. The IMI was computed and analyzed as described above. As seen in Fig. 1, the IMI tended to be inversely related to PLMS; that is, as the frequency of leg movements increased, the IMI decreased or remained stable. Statistical analysis revealed a significant difference between groups [F(1,132) = 5.96, p < 0.05] with group 2 patients displaying a longer IMI compared to group 1 patients (44.6 ± 1.6 seconds vs. 38.4 ± 1.2 seconds). In addition, a significant TON main effect [F(7,924) = 13.98, p < 0.01] and group by TON interaction [F(7,924) = 19.49, P < 0.01] were found. The latter interaction occurred due to the IMI increasing slightly as the night progressed compared to group 2 patients that remained fairly stable across the night (see Fig. 1). One possible explanation for the different TON patterns between groups is that group 2 patients had a higher number of leg movements during rapid eye movement (REM) sleep compared to group 1 patients. To test this possibility the number of leg movements RESULTS (collapsed across night) were compared in a three-way mixed-factors ANOVA. The between-groups factor TON patterns in PLMS activity was group (1 and 2) and the within (repeated)-group Figure 1 illustrates the averaged PLMS activity data factor was stage [nonrapid eye movement (NREM) and for groups 1 and 2 plotted in hourly blocks across the REM, NREM only, REM only] and TON (1-8). As night separately for nights 1 and 2. As seen, distinct can be seen in Fig. 2, there were no differences between TON patterns in PLMS activity are apparent: the first stages in the TON pattern (i.e. nonsignificant stage x pattern is characterized by a marked reduction in PLMS TON interaction). The main effect of stage was signifactivity across the night with the majority of the ac- icant [F(2,284) = 203.2, p < 0.01], i.e. REM movetivity (2:60%) occurring within the first 3 hours of ments (22.0) were significantly less (p < 0.01) than sleep. In contrast, the second pattern is characterized NREM movements (261.9) and all stages combined by a more stable pattern that increased across the first (283.9). half of the night and then decreased to a level similar Alternatively it could be argued that REM sleep disto the beginning of the sleep period with the majority tribution may account for the patterns. More specifiof the activity (2: 50%) occurring during the latter 4 cally, because REM sleep is known to inhibit PLMS, hours of sleep. These patterns are evidenced in each perhaps the two groups differed with regard to the disof the measures of PLMS activity with the frequency tribution of REM sleep. The latter is unlikely as PLMS of PLMS consistently greater than for the associated occurs predominantly in NREM sleep of which stage arousals. Furthermore, it can be seen in Fig. 1 that the 2 is relatively evenly distributed throughout sleep. Sleep, Vol, 15, No, 4,1992
308
W. J. CULPEPPER ET AL.
100 90
Group 1 - Night 1 (N=47)
(a)
•
o
80
F
o
PLMS EMG Arous(]ls EEG ArousQls
R 70
\;' Stage Shifts
E
.... IMI (sec)
Q
u
60 50
E 40 N C 30 Y 20
10 0 100 r---.---.---r---r---r---r---r---r--~r---.---r---r---r--'r--'r--'r--,~~100 Group 2 - Night 1 (N=25) Group 2 - Night 2 (N=25) 90 (e) t- (d) 90 F 80 R 70
•
o o
E
Q 60 U 50 E N 40 C 30
y 20 10
PLMS EMG Arousals EEG Arousals
\;' Stage Shifts .... IMI (sec)
&/• •
......
'l~.
~
~
'.
.
............... .
~ -
~
/-
A · . . . /.$ .,. ------{f- .• ' ::B=""~~ ;'" •
o-----o-~"~=~~"',..~
