SEARCH

SEARCH BY CITATION

Keywords:

  • human;
  • insomnia;
  • memory;
  • plasticity;
  • polysomnography;
  • sleep

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Study subjects
  6. Study design
  7. Mirror tracing task
  8. Visual and verbal memory task
  9. Cognitive test battery
  10. Sleep recordings
  11. Data analysis
  12. Results
  13. Discussion
  14. Disclosure Statement
  15. Acknowledgements
  16. References

It has been suggested that healthy sleep facilitates the consolidation of newly acquired memories and underlying brain plasticity. The authors tested the hypothesis that patients with primary insomnia (PI) would show deficits in sleep-related memory consolidation compared to good sleeper controls (GSC). The study used a four-group parallel design (= 86) to investigate the effects of 12 h of night-time, including polysomnographically monitored sleep (‘sleep condition’ in PI and GSC), versus 12 h of daytime wakefulness (‘wake condition’ in PI and GSC) on procedural (mirror tracing task) and declarative memory consolidation (visual and verbal learning task). Demographic characteristics and memory encoding did not differ between the groups at baseline. Polysomnography revealed a significantly disturbed sleep profile in PI compared to GSC in the sleep condition. Night-time periods including sleep in GSC were associated with (i) a significantly enhanced procedural and declarative verbal memory consolidation compared to equal periods of daytime wakefulness in GSC and (ii) a significantly enhanced procedural memory consolidation compared to equal periods of daytime wakefulness and night-time sleep in PI. Across retention intervals of daytime wakefulness, no differences between the experimental groups were observed. This pattern of results suggests that healthy sleep fosters the consolidation of new memories, and that this process is impaired for procedural memories in patients with PI. Future work is needed to investigate the impact of treatment on improving sleep and memory.


Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Study subjects
  6. Study design
  7. Mirror tracing task
  8. Visual and verbal memory task
  9. Cognitive test battery
  10. Sleep recordings
  11. Data analysis
  12. Results
  13. Discussion
  14. Disclosure Statement
  15. Acknowledgements
  16. References

Studies suggest that healthy sleep facilitates neural plasticity that is thought to underlie the consolidation of newly acquired and initially unstable memories (Diekelmann and Born, 2010; Stickgold, 2005). Periods of sleep after learning, relative to equal periods of wakefulness, have been shown to enhance both procedural memories for skills (e.g. Walker et al., 2002) and declarative memories for hippocampus-dependent, fact-based information (e.g. Gais and Born, 2004). Findings from animal (Wilson and McNaughton, 1994) and human studies (Maquet et al., 2000) have put forward the idea that, during sleep, novel memories are replayed and reprocessed in hippocampal–neocortical networks which might contribute to cortical long-term synaptic plasticity required for long-term memory storage (Ribeiro and Nicolelis, 2004). Complementary work has proposed that sleep-specific brain activity, such as sleep spindles (Schabus et al., 2004), ponto-geniculo-occipital waves (Datta et al., 2004), or electroencephalographic slow wave activity associated with synaptic downscaling (Tononi and Cirelli, 2006) might contribute to synaptic refinement and the strengthening of novel memory traces during sleep.

Compared to a compelling line of research in healthy humans and animals, only few studies have begun to investigate the impact of clinically disturbed sleep on memory consolidation. Preliminary findings on a subset of subjects (= 14) of the current study (= 86) suggest that overnight improvement in a procedural motor task (mirror tracing) might be impaired in patients with primary insomnia compared to good sleeper controls (GSC) (Nissen et al., 2006). One other study examined declarative memory consolidation during sleep (word-pair associate task) and found that this type of memory consolidation might be attenuated in patients with primary insomnia compared to GSC (Backhaus et al., 2006). However, both studies only assessed overnight memory consolidation and were limited by the lack of a wake comparison condition needed to demonstrate sleep-specific effects. The relevance of additional studies on sleep and memory in patients with sleep disorders has been addressed in the 2006 National Academy of Sciences Report, which states that ‘awareness should be increased about the importance of sleep to health, performance, and learning […]’ (Board on Health Sciences Policy, 2006).

The current study centred on memory consolidation during sleep and wakefulness in primary insomnia. Primary insomnia is a prevalent health problem affecting approximately 5% of the adult population (Ohayon, 2002). In addition to difficulties falling asleep, staying asleep or non-restorative sleep, daytime difficulties, such as memory dysfunction, are reported frequently by patients with primary insomnia. However, previous research that examined memory consolidation across periods of daytime wakefulness largely failed to provide objective evidence for subjective memory complaints (Fulda and Schulz, 2001).

This study was designed to investigate memory consolidation across equal periods of night-time including sleep and daytime wakefulness in patients with primary insomnia and GSC. We tested the hypothesis that, in healthy subjects, retention intervals containing sleep would foster the consolidation of procedural and declarative memories when compared to retention intervals containing only wakefulness, and that this effect of sleep would be impaired in patients with primary insomnia.

Methods

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Study subjects
  6. Study design
  7. Mirror tracing task
  8. Visual and verbal memory task
  9. Cognitive test battery
  10. Sleep recordings
  11. Data analysis
  12. Results
  13. Discussion
  14. Disclosure Statement
  15. Acknowledgements
  16. References

The study was conducted at the Sleep Laboratory at the Department of Psychiatry and Psychotherapy, University of Freiburg Medical Center. The clinical experiments conformed to the principles outlined by the Declaration of Helsinki and the procedures had been approved by the local ethic committee prior to the onset of the study.

Study subjects

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Study subjects
  6. Study design
  7. Mirror tracing task
  8. Visual and verbal memory task
  9. Cognitive test battery
  10. Sleep recordings
  11. Data analysis
  12. Results
  13. Discussion
  14. Disclosure Statement
  15. Acknowledgements
  16. References

Fifty-three healthy subjects and 33 patients with primary insomnia were examined (= 86). After complete description of the study, written informed consent was obtained from all participants. To ensure a representative sample, consecutive patients with insomnia scheduled for a sleep laboratory assessment as part of their clinical evaluation were invited to participate in the study. Insomnia patients met DSM-IV criteria for primary insomnia (American Psychiatric Association, 1994), as assessed by experienced sleep clinicians based on a clinical interview, standard physical examinations and sleep diaries.

Good sleeper comparison subjects were recruited from the community and paid for participation. Their good-sleeper status was ensured by clinical interviews and sleep diaries for 2 weeks. All participants underwent an extensive examination to rule out any comorbid physical or psychiatric disorder, including a Composite International Diagnostic Interview. The screening was performed by trained staff members under direct supervision of the sleep laboratory physicians.

All subjects were free of any medication for at least 2 weeks prior to the onset of the study and did not consume alcohol or caffeine during the study. All subjects were right-handed and non-smokers. A urine drug screening after the sleep laboratory night demonstrated that all participants were free of any benzodiazepines, barbiturates, amphetamines or opiates. Sleep diaries for 2 weeks ensured that the subjects’ usual sleep times approximated the imposed sleep schedule in the laboratory (±1 h). Sleep quality and subjective memory were assessed by using the Pittsburgh Sleep Quality Index (PSQI) (Buysse et al., 1989) and visual analogue scales.

Study design

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Study subjects
  6. Study design
  7. Mirror tracing task
  8. Visual and verbal memory task
  9. Cognitive test battery
  10. Sleep recordings
  11. Data analysis
  12. Results
  13. Discussion
  14. Disclosure Statement
  15. Acknowledgements
  16. References

The study used a four-group parallel design assessing the effects of 12 h of night-time including sleep versus 12 h of daytime wakefulness on memory consolidation in insomnia patients and GSC. Fifty-three healthy subjects and 33 insomnia patients were assigned to either a ‘sleep condition’ or ‘wake condition’. Procedural memory (mirror tracing task), declarative memory (visual and verbal learning task) and general cognitive performance were assessed either at 20:00 and 08:00 hours (‘sleep condition’) or at 08:00 and 20:00 hours (‘wake condition’), respectively, with equal retention intervals of 12 h between the learning and recall session in both conditions. The order of the tasks was balanced across subjects. Participants in the ‘sleep condition’ slept in the sleep laboratory with polysomnographic monitoring from 22:30 to 06:30 hours according to standard methods. During free periods of wakefulness, participants were instructed to follow their daily schedule. Continuous activity-records and actigraphy ensured that participants abstained from sleep or excessive mental/physical activity to keep levels of interference comparable.

Mirror tracing task

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Study subjects
  6. Study design
  7. Mirror tracing task
  8. Visual and verbal memory task
  9. Cognitive test battery
  10. Sleep recordings
  11. Data analysis
  12. Results
  13. Discussion
  14. Disclosure Statement
  15. Acknowledgements
  16. References

Procedural memory was assessed by using a mirror tracing task (Nissen et al., 2006; Plihal and Born, 1997). Participants were required to trace different line-drawn stimuli using a stylus with a light sensor that measured (1) draw time, (2) number of errors and (3) total error time. Error referred to the event when the stylus left the line. Subjects were asked to work as quickly and accurately as possible. Visual access to the stimuli was provided only indirectly via a mirror. In the learning session, mirror tracing a star was repeated until the participant reached a criterion of ≤15 errors to keep conditions comparable. Then, six line-drawn figures were presented. In the retrieval session, subjects traced the star once before the six figures of the learning session were presented. Initial learning was assessed by measuring the number of trials to criterion and measuring (1) mean draw time, (2) mean error count and (3) mean error time for mirror tracing the six figures. In the retrieval session, the same measures were assessed for tracing the six figures. Memory consolidation was calculated as the percentage of improvement from the learning to the retrieval session.

Visual and verbal memory task

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Study subjects
  6. Study design
  7. Mirror tracing task
  8. Visual and verbal memory task
  9. Cognitive test battery
  10. Sleep recordings
  11. Data analysis
  12. Results
  13. Discussion
  14. Disclosure Statement
  15. Acknowledgements
  16. References

To measure declarative memory, a standard visual and verbal memory task was used (Schellig and Schächtele, 2001). In this task, participants were required to memorize a line-drawn path on a map (visual learning) and information about the construction of a building provided in a text (verbal learning). After the retention interval, recall was measured without further presentation of the learning material. Memory consolidation was calculated as the percentage of retrieved items in the recall session referred to the learning session (retention rate, %).

Cognitive test battery

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Study subjects
  6. Study design
  7. Mirror tracing task
  8. Visual and verbal memory task
  9. Cognitive test battery
  10. Sleep recordings
  11. Data analysis
  12. Results
  13. Discussion
  14. Disclosure Statement
  15. Acknowledgements
  16. References

A standardized test battery was used in the learning and retrieval session to control for effects of short-term memory, psychomotor performance speed, alertness and divided attention (Zimmermann and Fimm, 2007).

Sleep recordings

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Study subjects
  6. Study design
  7. Mirror tracing task
  8. Visual and verbal memory task
  9. Cognitive test battery
  10. Sleep recordings
  11. Data analysis
  12. Results
  13. Discussion
  14. Disclosure Statement
  15. Acknowledgements
  16. References

Sleep laboratory recordings were performed from 22:30 to 06:30 hours and scored according to standard criteria (Rechtschaffen and Kales, 1968). The following variables of sleep continuity and architecture were assessed: sleep-onset latency Stage 1, defined as the period between ‘lights-out’ and the first 30-s epoch of Stage 1 sleep; sleep-onset latency Stage 2, defined as the period between ‘lights-out’ and the first 30-s epoch of Stage 2 sleep; sleep period time, defined as the period between sleep onset and the final awakening; sleep efficiency, defined as the ratio of total sleep time to time in bed × 100%; and time spent in waking and in sleep Stages 1, 2, slow wave sleep (SWS) (combined Stages 3 and 4) and rapid eye movement (REM) sleep, as a percentage of the sleep period time. REM sleep latency was defined as the period between sleep onset and the occurrence of the first 30-s epoch of REM sleep, including intermittent waking times (REM latency). REM density was calculated as the percentage (%) of the number of 3-s mini-epochs of REM sleep containing rapid eye movements referred to the number of 3-s mini-epochs of REM sleep containing no rapid eye movements.

Data analysis

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Study subjects
  6. Study design
  7. Mirror tracing task
  8. Visual and verbal memory task
  9. Cognitive test battery
  10. Sleep recordings
  11. Data analysis
  12. Results
  13. Discussion
  14. Disclosure Statement
  15. Acknowledgements
  16. References

Descriptive presentation of the data includes mean values and standard deviations. Independent t-tests were used to assess demographic and clinical differences between healthy subjects and insomnia patients at baseline. Two-factor multivariate analyses of variance (manovas) with the between-subject factors group (healthy subjects and insomnia patients) and condition (sleep and wakefulness) were performed to test for differences in memory encoding at baseline and memory consolidation (% change) from the encoding to the retrieval session. One-factor manovas with the factor group (healthy subjects and insomnia patients) were used to test for differences in sleep parameters and cognitive performance in the learning and recall condition. To quantify the difference between groups, partial ETA-squared (pETAsq) was computed, indicating small (0.01), medium (0.06) and large (0.12) effect sizes. To test for associations between sleep and memory parameters, Pearson’s correlation coefficients were calculated. The level of significance was set at < 0.05 (two-tailed).

Results

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Study subjects
  6. Study design
  7. Mirror tracing task
  8. Visual and verbal memory task
  9. Cognitive test battery
  10. Sleep recordings
  11. Data analysis
  12. Results
  13. Discussion
  14. Disclosure Statement
  15. Acknowledgements
  16. References

Participants

There were no significant differences in demographic characteristics between the four groups at baseline (Table 1). Patients with insomnia reported worse sleep (PSQI) than healthy subjects and scored worse on a visual analogue scale assessing subjective memory over the past 4 weeks. No significant differences were observed between the sleep and wake condition within the patient or comparison groups.

Table 1.   Study subjects
 Healthy controls (n = 53)Insomnia patients (n = 33)
Sleep condition (n = 36)Wake condition (n = 17)Sleep condition (n = 18)Wake condition (n = 15)
  1. Data are means ± standard deviations. Subjective memory refers to subjective memory capacity over the last 4 weeks as assessed by a visual analogue scale from 0 (poor) to 100 (excellent).

  2. IQ, intelligence quotient; PSQI, Pittsburgh Sleep Quality Index.

  3. Significant differences between insomnia patients and healthy subjects (sleep and wake condition) are indicated by superscript letters (t-tests for independent samples): *P<0.050; **P<0.010.

Male/female14/227/107/117/8
Age in years46.3 ± 4.947.5 ± 4.445.5 ± 4.547.1 ± 5.8
Age range in years40–5840–5538–5437–58
Full-scale IQ104.0 ± 11.5102.2 ± 9.0102.6 ± 13.4100.3 ± 10.1
Years in school11.5 ± 1.611.7 ± 1.811.6 ± 1.511.7 ± 1.8
Subjective memory55.1 ± 19.057.2 ± 25.043.3 ± 12.8*46.5 ± 20.3**
PSQI3.5 ± 1.62.1 ± 1.511.1 ± 8.2*12.4 ± 3.9**

Sleep

Polysomnographic results of insomnia patients and healthy subjects participating in the sleep condition are listed in Table 2. Patients with insomnia showed a significantly disturbed sleep continuity (reduced sleep period time, reduced sleep efficiency) and sleep architecture (increased waking time, reduced Stage 2 sleep, reduced REM sleep) compared to healthy subjects.

Table 2.   Polysomnographic parameters in the sleep condition
 Healthy controls (n = 36)Insomnia patients (n = 18)F10,43P-valuepETAq
  1. Data are presented as mean ± standard deviation. One-factorial multivariate analyses of variance with factor group.

  2. S1, Stage 1 sleep; S2, Stage 2 sleep; REM, rapid eye movement sleep, SWS, slow wave sleep; pETAsq, partial ETA-squared.

  3. Significant results are indicated by superscript letters: *< 0.050, **< 0.010, insomnia patients compared to healthy subjects.

Sleep latency (S1), min14.1 ± 12.321.1 ± 28.21.60.2100.03
Sleep latency (S2), min23.8 ± 18.730.3 ± 30.01.00.3310.02
Sleep period time, min451.7 ± 20.9433.1 ± 46.04.20.046*0.08
Sleep efficiency, %81.3 ± 10.170.3 ± 16.69.20.004**0.15
REM latency, min100.1 ± 42.6119.2 ± 62.61.70.1920.03
REM density, %26.0 ± 7.027.9 ± 12.20.50.4820.01
Sleep stage, as percentage of sleep period time
Waking, %13.6 ± 8.623.0 ± 12.510.40.002**0.17
Stage 1, %11.2 ± 4.912.5 ± 4.820.80.3580.02
Stage 2, %53.6 ± 7.546.1 ± 9.310.00.003**0.16
SWS, %3.7 ± 3.84.3 ± 5.20.20.6450.01
REM, %17.8 ± 5.014.0 ± 5.66.50.014*0.11

Memory

Table 3 displays absolute scores for memory performance. Insomnia patients and healthy subjects did not show any significant differences in initial learning performance at baseline (manova, > 0.1 for all main effects and pairwise comparisons).

Table 3.   Memory performance
 Healthy controls (n = 53)Insomnia patients (n = 33)
Sleep condition (n = 36)Wake condition (n = 17)Sleep condition (n = 18)Wake condition (n = 15)
LearningRecallLearningRecallLearningRecallLearningRecall
  1. Data are means ± standard deviations. Times are given in seconds. The values for verbal and visual memory represent the number of correctly retrieved items. No significant differences in procedural or declarative learning were observed at baseline.

Draw time110.1 ± 54.069.2 ± 29.892.0 ± 47.361.0 ± 21.991.4 ± 47.768.5 ± 30.7113.6 ± 47.387.3 ± 34.8
Error count19.6 ± 12.69.1 ± 8.114.1 ± 9.79.6 ± 8.114.0 ± 9.95.9 ± 4.812.9 ± 9.17.0 ± 6.3
Error time12.5 ± 10.08.1 ± 8.36.6 ± 6.75.3 ± 5.39.9 ± 7.65.0 ± 5.45.0 ± 5.13.1 ± 2.9
Verbal Memory14.9 ± 4.413.2 ± 4.614.1 ± 3.410.6 ± 3.613.7 ± 6.111.7 ± 6.613.7 ± 5.010.7 ± 5.6
Visual memory23.9 ± 4.721.0 ± 5.421.5 ± 5.216.7 ± 7.122.7 ± 5.818.8 ± 6.822.6 ± 5.719.0 ± 7.3

To analyse further memory consolidation from initial learning to retrieval, we calculated % improvements for procedural memory and % retention rates for declarative memory from the learning to the retrieval session (Table 4).

Table 4.   Memory consolidation from the learning to the retrieval test session
 Healthy controls (n = 53)Insomnia patients (n = 33)
Sleep condition (n = 36)Wake condition (n = 17)Sleep condition (n = 18)Wake condition (n = 15)
  1. Data are means ± standard deviations. Two-factor multivariate analyses of variance with factor group (healthy controls, insomnia patients) and factor condition (sleep, wake).

  2. Significant contrasts referred to the sleep condition in healthy controls are indicated by superscript letters: *< 0.05, **< 0.01, ***< 0.001, (*)= 0.05 to = 0.1, pETAsq = 0.05 to pETAsq = 0.07 (medium effect size).

% Improvement mirror tracing
 Draw time33.6 ± 15.622.4 ± 21.2*21.2 ± 17.2*19.0 ± 25.6*
 Error time38.5 ± 36.811.2 ± 57.2*52.6 ± 32.73.5 ± 119.5*
 Error count54.1 ± 23.037.7 ± 31.2*55.0 ± 21.919.5 ± 82.1(*)
% Retention rates visual and verbal learning task
 Verbal memory87.8 ± 10.674.6 ± 12.8***80.2 ± 23.9(*)74.9 ± 19.7**
 Visual memory88.3 ± 18.077.0 ± 27.1(*)81.8 ± 16.483.2 ± 23.8

Multivariate testing revealed a significant main effect for the factor condition (sleep/wakefulness; Wilks’ lambda: F5,78 = 3.1, = 0.012). The main effect for the factor group (Wilks’ lambda: F5,78 = 0.9, = 0.495) and the condition × group interaction (Wilks’ lambda: F5,78 = 1.1, = 0.356) did not reach statistical significance. Univariate tests indicated a significant condition effect for the parameter mirror tracing draw time (F5,78 = 4.2, = 0.03), error time (F5,78 = 8.0, = 0.006), error count (F5,78 = 5.7, = 0.019) and verbal memory (F5,78 = 10.5, = 0.002), but not visual memory (F5,78 = 0.8, = 0.386). To exclude that the condition effect for mirror tracing draw time was driven by a tendency towards slower mirror tracing performance in the evening in healthy subjects, we included the baseline performance as a covariate. This did not change the outcome of the analysis (F5,78 = 4.7, = 0.034).

To follow-up on the significant condition effect for procedural motor and declarative verbal memory, we explored this effect further in both groups separately. As indicated in Table 4, healthy subjects in the sleep compared to the wake condition demonstrated a significantly higher improvement in mirror tracing (draw time, error time and error count) and a significantly elevated verbal retention rate, consistent with the hypothesis that healthy sleep fosters the consolidation of procedural and declarative verbal memory consolidation. In contrast, no condition (sleep/wakefulness) effect was observed in insomnia patients.

To test directly our primary hypothesis that healthy sleep is associated with elevated memory consolidation compared to sleep in insomnia patients, we compared the sleep conditions between both groups. This analysis revealed that insomnia patients compared to healthy subjects showed a significantly reduced improvement in mirror tracing draw time and a non-significantly attenuated verbal retention rate across the sleep condition (Table 4). Across intervals of daytime wakefulness, healthy subjects and insomnia patients displayed very similar parameters of consolidation (> 0.4; pETAsq < 0.02 indicating small effect sizes).

The main findings on % improvements in mirror tracing draw time and % retention rates for declarative verbal memory are depicted in Figs 1 and 2, respectively.

image

Figure 1.  Procedural memory consolidation from the learning to the retrieval session. Bars represent means ± standard errors. Significant results are indicated by asterisks, *< 0.05.

Download figure to PowerPoint

image

Figure 2.  Declarative verbal memory consolidation from the learning to the retrieval session. Bars represent means ± standard errors. Significant results are indicated by asterisks, **< 0.01; ***< 0.001; (*)< 0.1.

Download figure to PowerPoint

Correlation analyses between sleep and memory parameters

The exploratory correlation analysis demonstrated a significant correlation between the sleep parameter REM density and % improvement in mirror tracing error time across both groups (Pearson’s correlation coefficient = 0.3, = 0.034), driven by a highly significant correlation within healthy subjects (Pearson’s correlation coefficient = 0.5, = 0.004; Fig. 3). No significant correlation was found within the insomnia group (Pearson’s correlation coefficient = 0.1, = 0.672). No other significant correlation between sleep and memory parameters was observed.

image

Figure 3.  Correlation between procedural memory consolidation across night-time periods containing sleep and rapid eye movement density in healthy subjects.

Download figure to PowerPoint

Additional neuropsychological assessments

Notably, insomnia patients and healthy subjects did not differ in any of the tasks in the standardized test battery in the learning or recall session (Table 5), indicating that the memory results were not driven by group differences in alertness, divided attention, short-term memory or psychomotor performance (> 0.1 for all tasks).

Table 5.   Cognitive performance
 Healthy controls (n = 53)Insomnia patients (n = 33)
Sleep condition (n = 36)Wake condition (n = 17)Sleep condition (n = 18)Wake condition (n = 15)
T1T2T1T2T1T2T1T2
  1. Data are means ± standard deviations. T1 and T2 refer to Time 1 and Time 2 of assessment. No significant differences between groups and conditions were observed (> 0.1).

Short-term memory7.8 ± 2.07.8 ± 2.27.5 ± 1.97.8 ± 2.17.5 ± 1.78.0 ± 2.67.7 ± 2.18.0 ± 2.4
Psychomotor speed31.6 ± 9.629.0 ± 9.030.6 ± 8.729.4 ± 8.131.1 ± 9.332.0 ± 9.730.0 ± 8.431.9 ± 10.6
Alertness255.6 ± 41.7250.4 ± 47.5257.5 ± 43.5252.3 ± 48.5259.5 ± 51.3260.5 ± 65.6257.4 ± 51.5259.1 ± 66.6
Divided attention680.6 ± 65.7685.9 ± 92.9685.6 ± 67.8687.9 ± 93.9684.3 ± 80.3693.1 ± 68.8682.9 ± 81.2690.1 ± 65.7

Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Study subjects
  6. Study design
  7. Mirror tracing task
  8. Visual and verbal memory task
  9. Cognitive test battery
  10. Sleep recordings
  11. Data analysis
  12. Results
  13. Discussion
  14. Disclosure Statement
  15. Acknowledgements
  16. References

The observed pattern of results is consistent with the concept that healthy sleep facilitates the consolidation of newly acquired procedural and declarative verbal memories. Evidence for this is that healthy subjects showed a significantly better improvement in mirror tracing (procedural memory) and a higher verbal retention rate (declarative memory) across a 12-h period that contained sleep than across a 12-h period that included only wakefulness.

Consistent with the initial hypothesis, patients with primary insomnia, in comparison to GSC, showed a significantly reduced improvement in mirror tracing across a 12-h period that included night-time sleep. Highly similar improvement rates in both groups in the wake condition suggest that the observed memory effects are specifically sleep-related. A limitation of the study is that this pattern of results, which can be interpreted as a differential effect of sleep on memory consolidation in both groups, was not reflected in a significant condition × group interaction, due perhaps to limited power. Against our prediction, no significant differences in declarative memory consolidation during sleep were observed between the experimental groups.

Over past years, studies in animals and humans have provided evidence that healthy sleep provides favourable conditions for memory consolidation and underlying brain plasticity (Diekelmann and Born, 2010; Stickgold, 2005). Our findings confirm these studies by demonstrating that, in healthy subjects, night-time periods containing sleep significantly enhance procedural and declarative memory consolidation in comparison to equal periods of daytime wakefulness. Our results extend this line of research in healthy subjects by corroborating a pilot report (Nissen et al., 2006) demonstrating that disturbed sleep is associated with decreased procedural memory consolidation in patients with primary insomnia. Moreover, our study addressed a significant limitation of this report by including a wake comparison condition. Highly similar procedural improvements across periods of daytime wakefulness in healthy subjects and insomnia patients indicate that patients with insomnia show specifically sleep-related and not mere time-related deficits in procedural memory consolidation.

Our observations in healthy subjects support the notion that both procedural and declarative memories benefit from sleep (Walker and Stickgold, 2004). However, we did not observe a significant disruption of sleep-related declarative memory consolidation in insomnia patients as described by Backhaus and colleagues (Backhaus et al., 2006). This disparity might be explained by studies proposing that sleep preferentially enhances associative declarative memories, as investigated in the word-pair associate task by Backhaus and colleagues, whereas the strengthening of on non-associative information—also relevant to everyday learning—as assessed in the present study, might be less dependent upon sleep (Rauchs et al., 2005; Robertson, 2009).

Regarding potential mechanisms, the observed positive correlation between REM density and motor skill consolidation during sleep in healthy subjects observed in the present sample might represent a polysomnographic correlate of preclinical findings demonstrating that neural activity related to the generation of rapid eye movements, such as ponto-geniculo-occipital waves, enhances experience-dependent cortical plasticity (Datta et al., 2004). The lack of correlation in insomnia patients would be consistent with the concept of altered REM sleep-related processes in these patients (Feige et al., 2008). Therefore, it seems plausible that alterations of REM sleep in patients with insomnia might contribute not only to subjective complaints of non-restorative sleep (Feige et al., 2008; Riemann et al., 2009) and emotional disturbances, as reported earlier (Greenberg and Pearlman, 1974), but also to impaired procedural memory consolidation, as demonstrated here. Importantly, it is notable that the correlation analysis has not been corrected for multiple-testing and has only an exploratory character. Of note, insomnia patients, compared to GSC, showed significantly reduced Stage 2 sleep—a sleep stage that has been implicated in motor skill consolidation (Walker et al., 2002). However, no significant correlation between Stage 2 sleep and procedural memory consolidation was observed in the current sample that would foster the concept that disruptions of this sleep stage might underlie reduced motor skill consolidation during sleep in the patient cohort.

Limitations of the study include that the present data are based on a single experimental night, without preceding adaptation to the sleep laboratory environment. This design may have contributed to a ‘first-night effect’ characterized by relatively poor sleep that might be pronounced in GSC (Agnew et al., 1966; Hauri and Olmstead, 1989). Thus, in contrast to the study by Backhaus and colleagues (Backhaus et al., 2006), we did not observe a significant reduction of SWS in the current sample of insomnia patients compared to healthy subjects. In addition to task characteristics, this divergence might have added to the lack of significant group differences in declarative memory consolidation in our study. Differences between healthy subjects and insomnia patients in the overnight consolidation of declarative memories might only emerge after adaptation to the sleep laboratory environment. Note that, with regard to mirror tracing, healthy subjects tended to perform worse in the evening and better in the morning with, at least in part, a tendency towards an opposed pattern in insomnia patients. This pattern could result from differences in homeostatic or circadian processes between insomnia patients and GSC. These alternative explanations, rather than differences in memory consolidation, might have contributed to the described results. However, to our knowledge there is no consistent evidence for alterations of general sleep homeostasis or circadian processes in patients with primary insomnia. Rather, non-significant differences between the experimental groups at baseline might reflect the large interindividual variance in absolute mirror tracing performance with differences emerging only in our main outcome parameter, memory consolidation from initial learning to retrieval.

Strengths of the study include a thorough control of demographic characteristics, full-scale IQ and levels of general cognitive performance. Another main strength of the study is the inclusion of a wake comparison condition that provides evidence for distinctly sleep-related memory effects. Clinically, the results of this study propose a novel, sleep-associated mechanism for memory complaints of patients with primary insomnia that are frequently reported in medical services and that could not be demonstrated by previous daytime studies (Fulda and Schulz, 2001).

Future studies are needed to determine whether, in addition to sleep alterations in primary insomnia, isolated insomnia symptoms in a variety of medical and mental disorders might also relate to memory dysfunction. From this perspective, sleep disturbances with high prevalence rates in patients with mental disorders, such as depression or schizophrenia may, independently from the mental disorder, result in disturbed memory functioning. Future work is needed to determine the impact of therapeutic interventions, such as pharmacotherapy or psychotherapy, on improving sleep and memory as two integral parts of health and functioning.

Disclosure Statement

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Study subjects
  6. Study design
  7. Mirror tracing task
  8. Visual and verbal memory task
  9. Cognitive test battery
  10. Sleep recordings
  11. Data analysis
  12. Results
  13. Discussion
  14. Disclosure Statement
  15. Acknowledgements
  16. References

Dr Nissen has received speaker honoraria from Sanofi-Aventis and Lundbeck. Dr Voderholzer has received speaker honoraria from Sanofi-Aventis, Lundbeck, Pfizer, Cephalon and Lilly. He has been principal investigator of an investigator-initiated trial sponsored by Lundbeck. Dr Riemann has served as an advisory board member of Sanofi-Aventis, Lundbeck, and GlaxoSmithKline. He has received speaker honoraria from Sanofi-Aventis, Lundbeck, GlaxoSmithKline and Servier and has received research support from Sanofi-Aventis, Omron, Organon, Takeda and Actelion. Dr Kloepfer, Dr Feige, H Piosczyk and Dr Spiegelhalder have no conflicts of interest to declare.

Acknowledgements

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Study subjects
  6. Study design
  7. Mirror tracing task
  8. Visual and verbal memory task
  9. Cognitive test battery
  10. Sleep recordings
  11. Data analysis
  12. Results
  13. Discussion
  14. Disclosure Statement
  15. Acknowledgements
  16. References

The authors wish to thank the technical staff and the Masters student Ute Goerke at the Department of Psychiatry and Psychotherapy, University of Freiburg Medical Center, for their help in conducting the study.

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Study subjects
  6. Study design
  7. Mirror tracing task
  8. Visual and verbal memory task
  9. Cognitive test battery
  10. Sleep recordings
  11. Data analysis
  12. Results
  13. Discussion
  14. Disclosure Statement
  15. Acknowledgements
  16. References