MSLT in Primary Insomnia: Stability and Relation to Nocturnal Sleep

Abstract

Study Objectives:

To assess the stability of the multiple sleep latency test (MSLT) in primary insomnia and its relation to total sleep time.

Design:

Randomized, double-blind, placebo controlled, clinical trial.

Setting:

Outpatient with sleep laboratory assessments in months 1 and 8 of treatment.

Participants:

Ninety-five primary insomniacs, 32-64 years old and 55 age- and sex-matched general population-based, representative controls.

Interventions:

After a screening nocturnal polysomnograms (NPSG) and MSLT the following day, participants with primary insomnia were randomized to take zolpidem 10 mg (n = 50) or placebo (n = 45) nightly for 12 months. During months 1 and 8, while taking their prescribed treatments, NPSGs and MSLTs the following day were conducted. A population-based sample served as controls and received a single NPSG followed by MSLT.

Results:

Mean daily sleep latency on the screening MSLT of insomniacs was normally distributed across the full range of MSLT scores and significantly higher than those of a population-based representative control sample (P < 0.006). The insomniacs with the highest screening MSLTs had the shortest screening total sleep times (P < 0.05). The MSLTs of insomniacs during treatment in study month 1 were correlated (r = 0.44, P < 0.001) with their month 8 MSLT. The mean MSLT score of the zolpidem group did not differ from that of the placebo group, and the stability within treatment groups also did not differ.

Conclusions:

These data support the hypothesis that some insomniacs show a reliable disorder of hyperarousal with increased wake drive both at night and during the day.

Citation:

Roehrs TA; Randall S; Harris E; Maan R; Roth T. MSLT in primary insomnia: stability and relation to nocturnal sleep. SLEEP 2011;34(12):1647-1652.

INTRODUCTION

Primary insomnia has been characterized as a disorder of hyperarousal. This hyperarousal has been shown in a number of physiological measures, including elevated levels of circulating catecholamines,¹ increased basal metabolic rate,² increased body temperature,³ altered heart rate,^4,5 elevated beta EEG frequency,^6,7 and cortical activation on functional neuroimaging.⁸ The most frequently cited finding suggesting that primary insomniacs are hyperaroused is the observation of elevated multiple sleep latency test (MSLT) results (i.e., mean daily sleep latencies) in primary insomnia relative to controls, although this finding is not without controversy.

Two of the earliest studies comparing MSLTs in patients with insomnia to control subjects failed to find significant MSLT differences,^9,10 although the one study with a large insomnia (n = 138) sample did report that 14% of the insomniacs “had no sleepiness” (i.e., MSLT = 20 min).¹⁰ Several other studies have similarly failed to find MSLT differences.^11–13 However, 2 of these studies had very small sample sizes,^11,12 and 2 used non-standard definitions of MSLT sleep onset.^12,13

The first study reporting increased MSLT scores in insomniacs relative to a control sample found a mean daily MSLT sleep latency of 14.7 min in insomniacs and 12.3 min in controls.¹⁴ A smaller study reported a mean daily sleep latency of 13.3 min in insomniacs and 9.5 min in controls.² A more recent larger study reported a significant difference in insomniacs and controls with mean daily sleep latencies of 10.3 vs 8.6 min.¹⁵

What is evident in the studies above reporting significant MSLT differences between insomniacs and controls are the differences among the studies in the control sample MSLT mean sleep latencies, which range from 8.6 min to 12.3 min. All these studies used samples of convenience, recruiting their control samples through newspaper advertisements for normal sleepers. To better understand whether the MSLT is elevated in insomniacs it would be valuable to compare insomniacs to a population-based, representative sample of healthy non-insomniacs.

The studies of MSLT in insomnia also differ in the definition of insomnia used. Some of the above cited studies defined insomnia based on a single complaint only, others used the standard clinical diagnostic criteria, and others included nocturnal polysomnographic (NPSG) entry criteria. How nocturnal sleep time relates to MSLT in insomnia is an important theoretical question. In studies of healthy normals, short MSLT mean daily sleep latencies are associated with diary-reported short total sleep times and high laboratory sleep efficiencies the night before the MSLT.¹⁶ These results suggest that the less the habitual nocturnal sleep, the greater the degree of daytime sleepiness. The question arises as to how sleep time in insomniacs relates to their MSLTs the following day.

Finally, the studies of MSLT in insomnia report relatively large standard deviations. The wide variability among insomniacs may reflect high test-retest variability in a given individual's MSLT. No study has assessed the repeatability of MSLT results in insomniacs across time. Will an insomniac with an elevated MSLT show the same elevation across time? In this paper, we report MSLT results of a large sample of insomniacs participating in a chronic pharmacological treatment trial; compare these results to a representative population control sample; evaluate the relation of MSLT to NPSG sleep time; and importantly, assess the stability of the MSLT in insomnia.

METHODS

Participants

Persons 21-70 y old with difficulty falling asleep and/or staying asleep were recruited through newspaper advertisements. Ninety-five participants (42 men, 53 women), aged 32-64 years, meeting Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition, Text Revision (DSM-IV-TR) criteria for primary insomnia were recruited for a clinical trial of the efficacy and abuse liability of chronic hypnotic use (see Table 1 for insomnia sample demographics). All were in good physical and psychiatric health based on medical, psychiatric, drug use history, and physical examination and screening blood and urine laboratory analyses (see below).

Table 1.

Demographics, sleep, and drug use history of insomnia groups

	Placebo	Zolpidem
N	45	50
Average Age (range)	49.44 (30-70)	49.58 (23-68)
Gender
Males	16	26
Females	29	24
Insomnia Measures
Reported Sleep Time (h) ± SD; Median	5.58 ± 1.04; 5.50	5.17 ± 1.22; 5.25
NPSG Sleep Time (h) mean ± SD; median	5.90 ± 0.74; 6.02	6.02 ± 0.78; 6.12
SL NPSG (min) ± SD	44.26 ± 38.35	35.46 ± 29.32
median; range	33; 2–169	26; 5–141
WASO NPSG (min) ± SD	98.44 ± 40.67	95.15 ± 42.26
median; range	91.5; 22–185.5	92.0; 24.5–218.5
Age of insomnia onset median; range	38.60 40; 10.5–65	36.46 38.5; 7–58
Duration of insomnia median; range	10.93 7; 1–31	12.92 8; 1–54
Self Reported Alcohol and Drug Consumption
Alcohol (drinks/week: N)
0-1	31	31
2-6	6	16
7-14	7	3
16	1	0
Caffeinated Beverages (drinks/week: N)
0-1	9	13
2-6	7	16
7-14	16	12
16 or more	13	9
Previous Illicit Drug History
No drug history	31	41
Marijuana use ≥ 2 y ago	3	5
Marijuana use > 10 y ago	10	4
Cocaine use 20 y ago	1	0
Current Nicotine Usage (Cigarettes smoked per day)
Non smokers	37	41
1-2	3	3
4-5	2	4
> 6	3	2

The control population was selected from the Southeastern Michigan Sleep Survey database which contains a randomly drawn population representative sample from the Metropolitan Detroit area.¹⁶ The controls were selected to be similar to the age and sex distribution of the primary insomnia sample. The control sample included 27 men and 28 women between the ages of 22 and 65 years (see Table 2 for control sample demographics). All were in good physical and psychiatric health based on the brief self-reported screening questionnaires described below. The screening NPSGs and MSLTs of the 2 samples were conducted in the same fashion as described below. The Institutional Review Board of the Henry Ford Health System approved both study protocols. All participants provided informed consent and were paid for participation.

Table 2.

Demographics and sleep of the control group

N	55
Average Age (range)	48.09 (22-65)
Gender
Males	27
Females	28
Reported sleep time (h) mean + SD; median	6.90 ± 1.22; 6.90
NPSG sleep time (h) mean + SD; median	7.12 ± 0.74; 7.33

Study Design

To assess the presence of MSLT elevations in primary insomnia, the screening MSLT distribution of the insomnia sample was compared to the MSLT distribution of the control sample. Then the insomnia subjects entered a treatment trial, which was a mixed design, double-blind, placebo-controlled, study with a between group comparison of insomniacs randomly assigned to use zolpidem 10 mg (5 mg for those > 60 y) or placebo nightly for 12 months. After the screening NPSG and MSLT, insomnia subjects received a single NPSG and MSLT on the first night of zolpidem or placebo administration and again after 8 months of nightly use of their assigned medication. Insomniacs took their assigned medications on the NPSG night (30 min before bedtime) before the month 1 and month 8 MSLT. To assess the stability of MSLT and its relation to sleep time, the second goal of this study, the month 1 and month 8 NPSGs and MSLTs of the insomnia sample were compared.

Procedures

General health and psychiatric screening of insomniacs

Respondents to advertisements for persons “with difficulty falling asleep and/or staying asleep” were interviewed by telephone regarding their insomnia, general health, past and current psychiatric, alcoholism, and drug abuse histories. Those passing the telephone screen were scheduled for a clinic visit, during which they gave informed consent, provided a medical and drug-taking history, including prescribed medications, legal and/or illegal recreational drugs, underwent a physical examination, and provided blood and urine samples. Those with any acute or unstable illness, including conditions making it unsafe for the person to participate, conditions with a potential to affect sleep (i.e., acute pain, respiratory disorders), and conditions which could interact with the pharmacokinetics or pharmacodynamics (e.g., moderate to severe liver disease, heavy smoking) of zolpidem were excluded, as were those with chronic illnesses including renal failure, liver disease, seizures, and dementing illnesses. Participants with current psychiatric disorders, anxiety, depression, and schizophrenia disorders, as identified by the Structured Clinical Interview for DSM-IV-TR (SCID) were also excluded.

The blood and urine panel also included testing for drugs, including opiates, benzodiazepines, and stimulants. Participants with a history of substance (drug or alcohol) use disorders or current use of central nervous system acting medications at screening were excluded. Participants reporting any use of illegal drugs within the past 2 years also were excluded. Participants who reported consuming > 14 standard (1 oz) alcoholic drinks per week (with 1 exception, a person consuming 16 drinks per week in the placebo group), caffeine consumption > 300 mg/day, and smoking during the night (23:00-07:00) were excluded (see Table 1 for drug use histories).

Sleep Disorders Screen

Each participant underwent a sleep-wake history, sleep disorders evaluation, and NPSG. Each qualified for a DSM-IV-TR diagnosis of primary insomnia. As part of the sleep-wake history, participants completed a 2-week sleep diary, which was used to determine their habitual bedtime and the screening NPSG bedtime. For the screening NPSG bedtime, the midpoint of their bedtime reported on the 2-week sleep diary was determined, and 4 h were added to each side of the diary-reported bedtime midpoint to create an 8-h bedtime that would not disrupt circadian rhythms.

After the screening physical and laboratory tests, the participants underwent the screening 8-h NPSG.¹⁷ In addition to the clinical DSM-IV-TR primary insomnia diagnosis, all insomnia participants were also required to show sleep efficiencies ≤ 85% (total sleep time/time in bed) and no other primary sleep disorders on the screening NPSG. Participants with respiratory disturbances (apnea hypopnea index [AHI] > 10) or leg movements (periodic limb movement arousal index [PLMAI > 5]) were excluded from the study.^18,19 There were no MSLT criteria applied.

The standard Rechtschaffen and Kales methods for recording of sleep were used.¹⁷ The NPSGs included standard central (C3-A2) and occipital (Oz-A2) channels for electroencephalogram (EEG), bilateral horizontal electrooculograms (EOG), submental electromyogram (EMG), and electrocardiogram (ECG) recorded with a V5 lead. In addition, on the screening night airflow was monitored with oral and nasal thermistors, and leg movements were monitored with electrodes placed over the left tibialis muscles; respiration and tibialis EMG recordings were scored for apnea and leg movement events; and event frequencies were tabulated.^18,19 After screening, subsequent NPSGs excluded airflow and leg monitoring. All NPSGs were scored in 30-sec epochs according to the standards of Rechtschaffen and Kales.¹⁷ Scorers maintained 90% interrater reliability.

The MSLT was performed according to the standard protocol,²⁰ the first test occurring between 1.5-3 h after arising and then every 2 h thereafter for a total of 4 tests. Participants lay down in a bed in a quiet and dark room with the instruction to go to sleep. On each test participants remained in bed for 20 min if sleep onset did not occur. For both insomniacs and population controls, the subjects remained in bed for 15 min after sleep onset defined as minutes from lights out to the first 30-sec epoch scored as sleep. The MSLT was scored by experienced scorers who are unaware of the participants' study group. The dependent measure was average daily sleep latency in minutes averaged over the 4 tests.

Control Population Methods

The control population was drawn from the Southeastern Michigan Sleep Survey database which is a large epidemiologic/laboratory study of excessive daytime sleepiness in the general population.¹⁶ Briefly, 3,283 respondents, selected using a random digit dial procedure yielding a sample representative of the population of southeastern Michigan, completed a 20-min telephone survey querying sleep and health habits and general information regarding medical, psychiatric status and medication use. From among the 3,283 telephone survey respondents, 399 were randomly chosen for a single night of NPSG followed by MSLT. On the laboratory night the Global Sleep Assessment Questionnaire and the Diagnostic Interview Schedule for DSM-IV were completed. The population controls used in this study were selected from among the 399 laboratory study participants. As in the insomnia sample, the control sample completed a 2-week sleep diary, which as described above, was used to establish their NPSG bedtimes. The sleep diary was not used as an exclusionary screen in either sample. The control subjects were selected for no psychiatric or medical diseases, daily alcohol use, drug dependency, restless legs complaints, or insomnia (using the DSM-IV Research Diagnostic Criteria questions), shift or night work, and they showed no primary sleep disorders on their NPSG. The NPSGs and MSLTs of the control sample were conducted in the same manner as that of the screening NPSGs and MSLTs in the primary insomnia sample as described above. The same primary sleep disorder exclusionary criteria of the insomnia sample were applied to the control sample. As with the insomnia sample, no MSLT criteria were applied, but unlike the insomnia sample, no NPSG sleep efficiency criteria were applied. The self-reported sleep times and NPSG sleep times of the control sample are included in Table 2.

Statistical Analyses

A simple t-test was used to compare the mean MSLT scores (average latency in min over the 4 tests) of insomniacs to controls. The distribution of MSLT scores among the insomniacs was divided into quartiles, as was the distribution of MSLT scores among the controls. The screening total sleep times of MSLT quartiles were compared using between-group analyses of variance (ANOVAs). Pearson product correlations were conducted comparing the month 1 to month 8 MSLT scores. Multivariate analyses of variance (MANOVAs) were used to compare MSLT scores between the placebo and zolpidem groups in month 1 and month 8.

RESULTS

The MSLT scores of the insomniacs on their screening day ranged from 2-20 min and were normally distributed (Shapio-Wilk test statistic: 0.95, P < 0.001). The mean MSLT score of insomniacs was 13.2 ± 4.65 min, and the median was 13.6 min. Figure 1 presents the box plots of the insomnia MSLT distribution in comparison to the non-insomniac control MSLT distribution. The MSLT mean of the control sample was 11.0 ± 4.93 min, which was significantly shorter than that of the insomniacs (t = 2.79, P < 0.006). There were significantly more insomniacs having MSLT scores > 11 min than in the control population (70% vs 50%; X² = 5.17, P < 0.02). Among the insomniacs, the mean screening MSLT score of those randomized to the placebo group did not differ from those assigned to the zolpidem group (14.0 ± 4.37 vs 12.8 ± 4.91).

Figure 1.

Box plots comparing the distribution of MSLT mean daily sleep latency between the population-representative, non-insomniac controls and primary insomniacs. The box reflects the 25^th to 75^th percentiles, the line in the middle of each box the median, and the whispers the 95^th confidence intervals (CI). Separate data points beyond the CI are outliers.

The distribution of screening MSLT scores among insomniacs was divided into quartiles; mean MSLT of the highest quartile was 18.2 ± 1.5 min, and in the lowest quartile it was 7.3 ± 2.6 min. The 2 intermediate quartiles were collapsed, and the mean was 13.6 ± 1.4 min. Total sleep time (TST) on the screening night, immediately preceding the MSLT, was significantly shorter in the highest MSLT quartile (344.4 ± 51.6 min) than that of the lowest quartile (374.4 ± 28.6 min) (F = 3.15, P < 0.05). The TST of the 2 intermediate quartiles was 350 ± 45.9 min and not different than either of the 2 extremes. A similar division of the MSLT distribution of the control population yielded a mean MSLT in the highest quartile of 17.1 ± 2.0 min, 10.4 ± 1.9 min in the intermediate 2 quartiles, and 4.4 ± 1.5 in the lowest quartile. In Figure 2, the TST of the control population at each quartile is compared to that of the insomnia population at each quartile. A 2-factor ANOVA comparing the TST of the insomnia and control populations as a function of MSLT quartile showed a main effect of group (F = 91.75, P < 0.001) and a main effect of quartile (F = 6.65, P = 0.002) and no interaction. Within the control population, the TST of the highest MSLT quartile was shorter than that of the lowest quartile (F = 3.28, P < 0.05). At each quartile, the TST of the insomnia group was shorter than that of the control population (P < 0.001).

Figure 2.

Screening total sleep time (TST) of insomnia and control populations at each of the MSLT quartiles. TST was shorter in insomniacs compared to controls (F = 91,75 P < 0.001) and at the highest MSLT quartile relative to the lowest (F = 6.65, < 0.002) with no interaction.

The relation of MSLT mean daily sleep latency in month 1 to that of month 8 is presented in Figure 3. The correlation between month 1 and month 8 across all subjects (N = 95) was r = 0.44 (P < 0.001); the correlation for subjects in the placebo group (n = 45) was r = 0.43 (P < 0.01), and for subjects in the zolpidem group (n = 50) it was r = 0.49 (P < 0.001). The percentage of insomniacs whose MSLT score was greater than the control population mean of 11 min in month 1 and remained higher in month 8 was 87%, while those whose month 1 MSLT score was lower and remained lower in month 8 was 64% (X² = 3.56, P < 0.059). Table 3 presents the mean MSLT score of the placebo and zolpidem groups in month 1 and 8. There were no differences in MSLT scores between the drug groups or between month 1 and 8.

Figure 3.

The mean daily MSLT sleep latency (min) for both placebo (open circles) and zolpidem (filled circles) groups in month 1 plotted against that in month 8 (r = 0.44, P < 0.001).

Table 3.

Mean daily MSLT sleep latency (min)

	Month 1	Month 8
Placebo	13.6 ± 5.41	12.9 ± 5.37
Zolpidem	12.6 ± 5.09	12.9 ± 4.85

DISCUSSION

These are the first data to compare the MSLT of a large sample of primary insomniacs to that of a healthy, non-insomniac, population-based representative control sample and to have tested the stability of the MSLT in insomniacs over an extended interval of time. Relative to the population, the mean MSLT score of the insomniacs is higher than that of the controls. However, the distribution of MSLT scores in insomniacs is a normal distribution and covers the whole MSLT range, 2-20 min. In other words some insomniacs have high MSLT scores, while others do not.

It should be acknowledged that the screening procedures used in obtaining the control sample and the insomnia sample differed. Importantly, the NPSG and MSLT procedures on which the two samples were compared did not differ. The insomnia sample was screened extensively in both telephone and face-to-face interviews to select a clean primary insomnia sample. The screening of the control sample was more limited as its purpose was to generate a population representative sample. These differences may have biased the results, but it is not clear as to how.

The MSLT in insomniacs did not show the same relation with duration of sleep the previous night as seen in non-insomniacs. In insomniacs TST was consistently shorter than controls regardless of MSLT quartile. These findings suggest that the MSLT in insomniacs reflects a basic characteristic of the pathophysiology of insomnia, rather than the quantity and quality of nocturnal sleep as is seen in healthy controls or in patients with obstructive sleep apnea syndrome.¹⁶ Insomniacs have been described as being “hyperaroused,” as reflected in various physiological measures. However, to date only a single study has shown “hyperarousal” on concurrent physiological measures, those being increased metabolic rates in conjunction with elevated MSLTs.²

An important observation in this large sample of insomniacs is that there are “sleepy” insomniacs and the question arises as to how “sleepy” insomniacs differ from “hyperaroused” insomniacs. These present data would indicate that “sleepy” insomniacs sleep significantly longer than “hyperaroused” insomniacs (6.2 h vs 5.7 h). This may indicate a difference in the severity of their insomnia with the “sleepy” insomniacs expressing a less severe insomnia than that of the “hyperaroused” insomniacs. It has been suggested that comorbid insomniacs differ from primary insomniacs in that they are “sleepy” insomniacs.²² This has been reported in patients with insomnia and periodic limb movements and those with insomnia and rheumatoid arthritis. Effective treatment with hypnotics that increased nocturnal sleep time also increased MSLT scores in these comorbid insomniacs.²² It should be noted that this study was carried out exclusively in primary insomniacs and among primary insomniacs some are “sleepy” and the majority are “hyperaroused.” And as discussed below effective hypnotic treatment did not alter their MSLT.

As to the daytime expression of insomnia, one wonders whether “sleepy” and “hyperaroused” primary insomniacs differ on other indices of daytime physiological arousal beyond the MSLT. The previously cited study showing increased metabolic rate in insomniacs had too small a sample to make comparisons among the insomniacs based on those with high versus low MSLTs.² Analyses of urinary catecholamines and cortisol in this large sample of primary insomniacs are currently being conducted.

The MSLT of insomniacs was a stable finding. While the correlation of month 1 to month 8 MSLT scores was significant, it was modest with regard to the amount of variability predicted. Critically, relative to the control population with a mean MSLT of 11 min, for the vast majority (87%) of those insomniacs whose mean MSLT was greater than 11 min in month 1, it remained so in month 8. What is of interest is that some of these insomniacs received placebo nightly for the 8 months while others received zolpidem. As seen in Table 3, the MSLT did not differ between the placebo and zolpidem groups and did not change over the 8 months. This finding again supports the notion that MSLT in insomniacs may not reflect the quantity or quality of nocturnal sleep as it does in non-insomniacs.

It is also interesting to note that eight months of treatment did not alter the daytime expression in these insomniacs of “hyperarousal” as defined by MSLT. Active drug relative to placebo did improve nocturnal sleep both in month 1 and month 8, which has been presented in abstract form.²³ Relative to the placebo group, sleep time was still increased by 48 min in month 8 in the zolpidem treated group. Hypnotic treatment is often described as a symptomatic treatment and these data may reflect that description. To the extent that daytime “hyperarousal” reflects the underlying pathophysiology of insomnia, it may not have been altered with 8 months of improved nocturnal sleep.

On the other hand, the MSLT is a nonspecific test; it merely measures the time to fall asleep when given the opportunity. Short MSLT latencies can be due to narcolepsy, the sleep fragmentation of apnea or PLMs, a sedating drug, or chronic insufficient sleep. In the same way, unusually long MSLT latencies could be due to “hyperarousal” in insomnia, a vigorous walk before each MSLT test, or full alertness in a healthy normal. Earlier we noted the importance of concurrent indices of physiological arousal beyond the MSLT in differentiating “sleepy” and “hyperaroused” insomniacs. The same case can be made for understanding high MSLT scores in treated insomniacs. Do the high scores reflect continued “hyperarousal” or full alertness after 8 months of improved sleep? Further study utilizing multiple measures of “hyperarousal” is necessary to answer that question.

DISCLOSURE STATEMENT

This was not an industry supported study. Dr. Roehrs has consulted for Elan, Sanofi- Aventis, and Sepracor and has participated in speaking engagements for Sanofi-Aventis and Sepracor. Dr. Roth has received research support from Aventis, Cephalon, GlaxoSmithKline, Neurocrine, Pfizer, Sanofi-Aventis, Schering Plough, Sepracor, Somaxon, Syrex, Takeda, TransOral, Wyeth, XenoPort; has consulted for Abbott, Accadia, Acoglix, Actelion, Alchemers, Alza, Ancil, Arena, AstraZeneca, Aventis, BMS, Cephalon, Cypress, Dove, Elan, Eli Lilly, Evotec, Forest, GlaxoSmithKline, Hypnion, Johnson and Johnson, King, Ludbeck, McNeil, MediciNova, Merck, Neurim, Neurocrine, Neurogen, Novartis, Orexo, Organon, Orginer, Prestwick, Procter and Gamble, Pfizer, Purdue, Resteva, Roche, Sanofi, Schering Plough, Sepracor, Servier, Shire, Somaxon, Syrex, Takeda, TransOral, Vanda, Vivometrics, Wyeth, Yamanuchi, and XenoPort and has participated in speaking engagements for Sanofi, Cephalon, and Takeda. The other authors have indicated no financial conflicts of interest.