View More View Less
  • 1 Centre for Cognitive Neuroscience (CCNS), University of Salzburg, Austria
  • 2 University of Salzburg, Austria
Open access

Today, there is little doubt concerning the significance of sleep for memory consolidation. Some studies have suggested, however, that overnight memory consolidation as well as their underpinning neural mechanisms might be modulated by general cognitive abilities. In this paper, we used a more specific trait measure of declarative word-pair encoding efficiency, namely “baseline memory performance (BMP).” We explored its relation to consolidation and stabilization of declarative memories overnight as well as its relationship to sleep mechanisms. We included healthy subjects and insomnia patients from two studies with slightly differing demands on declarative memory. In the first study, an insomnia sample (N = 21) performed a declarative word-pair association task with pre- and post-sleep retrieval sessions and recorded 8 hr of nocturnal sleep following learning. In the second study, insomnia (N = 24) as well as sex-and-age-matched control (N = 29) subjects underwent a similar task but with an additional interference and time delay manipulation. Based on their encoding efficiency in the evening, all three study samples were split into good (BMP+) and moderate (BMP−) learners. Although not each subsample reached statistical significance, we observed – across all three samples – a pattern of only BMP− forgetting overnight and BMP+ showing enhanced activity and density of sleep spindles. Our findings suggest that – independent of the exact study design and subjective sleep complaint – exclusively participants with high BMP seem to be able to eliminate forgetting over sleep which may be related to stronger “offline” replay.

Abstract

Today, there is little doubt concerning the significance of sleep for memory consolidation. Some studies have suggested, however, that overnight memory consolidation as well as their underpinning neural mechanisms might be modulated by general cognitive abilities. In this paper, we used a more specific trait measure of declarative word-pair encoding efficiency, namely “baseline memory performance (BMP).” We explored its relation to consolidation and stabilization of declarative memories overnight as well as its relationship to sleep mechanisms. We included healthy subjects and insomnia patients from two studies with slightly differing demands on declarative memory. In the first study, an insomnia sample (N = 21) performed a declarative word-pair association task with pre- and post-sleep retrieval sessions and recorded 8 hr of nocturnal sleep following learning. In the second study, insomnia (N = 24) as well as sex-and-age-matched control (N = 29) subjects underwent a similar task but with an additional interference and time delay manipulation. Based on their encoding efficiency in the evening, all three study samples were split into good (BMP+) and moderate (BMP−) learners. Although not each subsample reached statistical significance, we observed – across all three samples – a pattern of only BMP− forgetting overnight and BMP+ showing enhanced activity and density of sleep spindles. Our findings suggest that – independent of the exact study design and subjective sleep complaint – exclusively participants with high BMP seem to be able to eliminate forgetting over sleep which may be related to stronger “offline” replay.

Highlights

  1. Overnight declarative memory consolidation depends on encoding efficiency
  2. “Bad learning” in the evening predicts overnight forgetting of facts
  3. Good learners show more (fast) sleep spindles
  4. Better fact learning coincides with higher amounts of slow-wave sleep in insomnia

Introduction

Sleep has been repeatedly linked to the consolidation of newly acquired information (Diekelmann and Born, 2010). On a systemic level, it is believed that initially new information is stored in two parallel memory systems: the hippocampus, a fast learning but short-term memory system and the neocortex, a slow learning but long-term memory system (McClelland, Mcnaughton, and O’Reilly, 1995). After memory acquisition, new memories become spontaneously reactivated as soon as the brain enters a state of quiet wakefulness or sleep. It has been shown that such reactivations mainly occur in coincidence with sharp wave–ripple complexes within the hippocampal formation and might be crucial for memory consolidation (Ego-Stengel and Wilson, 2010; Girardeau, Benchenane, Wiener, Buzsáki, and Zugaro, 2009). Moreover, hippocampal reactivations have been found to be temporally coupled to thalamocortically generated sleep spindles (SS) (Siapas and Wilson, 1998) and the up-state of slow oscillations (SO) during deep sleep (Clemens et al., 2007; Mölle, Yeshenko, Marshall, Sara, and Born, 2006). Whereas SS are proposed to support synaptic plasticity (Rosanova and Ulrich, 2005), SO orchestrate hippocampal reactivations and SS into time frames, where most parts of the cortex are depolarized and are thus more susceptible to plastic changes (Diekelmann and Born, 2010). By this means, according to the theory, new cortical memories get gradually strengthened and gradually independent of the short-term representation within the hippocampus.

In addition to their role in “offline” consolidation, SS are supposed to reflect the efficiency of thalamocortical connectivity and are thereby linked to different trait-like abilities, such as memory aptitude and logical reasoning (Bodizs et al., 2005; Fenn and Hambrick, 2011; for review see Fogel and Smith, 2011; Nader and Smith, 2001; Schabus et al., 2006; Schabus et al., 2008; Tucker and Fishbein, 2009).

In this paper, we sought to explore the influence of baseline memory performance (BMP) – i.e., a measure of encoding efficiency of new declarative material – on forgetting or the degree of overnight memory consolidation, as well as its influence on sleep mechanisms associated with the “offline” consolidation. For reliably characterizing the influence of BMP on overnight memory change (OMC), we included healthy subjects and insomnia patients from two conducted studies with slightly differing demands on memory (overnight forgetting and susceptibility to interference).

We hypothesize that also subjects with higher BMP have more effective thalamocortical and intracortical connectivity that supports the more efficient encoding process. This in consequence allows stronger reactivation during sleep, which ultimately results in better overnight memory consolidation.

Methods

In this paper, we analyze data from three samples of participants acquired in two separate studies, that is, one study on time dependent forgetting with insomnia subjects (study 1) and one study on memory susceptibility with insomnia patients as well as age-matched healthy controls (study 2).

Subjects

Participants for both studies were recruited at the University of Salzburg and using public advertisements. Study 1 and study 2 were a part of a larger study investigating the effects of neurofeedback on primary insomnia over multiple neurofeedback sessions and nights (study 1, see Schabus et al., 2013; study 2, Griessenberger et al., 2013). All of the subjects reported smoking less than five cigarettes a day and were free of any medication (other than oral contraceptives) for at least 2 weeks prior to the onset of the study. During the pre-screening session preceding the screening night by around a week as well as during screening night session, subjects went through a number of psychometric questionnaires to monitor their sleep quality, anxiety, depression, personality, and quality of life. Clinical screening interview served as an additional source of information about participant’s medications, disorders, etc. At study intake, all participants provided written informed consent, and at study completion, they got monetary compensation. During the whole study, subjects’ sleep habits were monitored with sleep diaries and wrist actigraphy.

The inclusion criterion for our insomnia groups was poor subjective quality of sleep. Furthermore, these participants were diagnosed as insomnia patients when their global score on the Pittsburgh Sleep Quality Index Questionnaire (PSQI; Buysse, Reynolds, Monk, Berman, and Kupfer, 1989) exceeded 5 points and when their insomnia complaint showed some independence from the temporal course of other mental or psychiatric disorders (verified by the Structured Interview for Sleep Disorders According to DSM-III-R; Schramm et al., 1993), according to the research diagnostic criteria of Edinger et al. (2004). Additional questionnaires for depression, anxiety, and quality of life were given in both studies [BDI (Beck, Steer, and Brown, 1996), BAI (Beck, Epstein, Brown, and Steer, 1988)/STAI (Spielberger, 1983), and WHOQOL (The WHOQOL Group, 1998)]. Participants who reported no subjective sleep quality complaint and who had a global PSQI score below 5 served as control subjects.

In study 1, we analyzed data from 23 subjects (6 males) diagnosed with primary insomnia, aged between 19 and 48 years (M = 34.2, SD = 10.4). One participant was excluded from analyses as he was unable to learn the memory task (correctly recalled items before sleep was only 5%). PSQI scores in this sample varied between 6 and 14 points (M = 10.9, SD = 2.0). In addition, a test verifying memory ability (Wechsler Memory Scale-revised, WMS-R; Wechsler, 1987) was performed at study intake.

In study 2, we analyzed data from a group of insomnia subjects as well as a healthy control group. During data acquisition, the two samples were balanced with respect to age and sex. The final sample of insomnia patients consisted of 24 subjects (9 males), aged between 23 and 58 years (M = 36.8, SD = 11.2). PSQI scores in this sample varied between 6 and 19 points (M = 10.6, SD = 3.3). The final sample of healthy controls consisted of 29 subjects (14 males), aged between 20 and 55 years (M = 35.8, SD = 10.0). PSQI score in this sample varied between 0 and 5 points (M = 3.0, SD = 1.3). The χ2 tests for independence revealed no statistically significant relationship between gender and sample (healthy control, insomnia patient): χ2(1) = .621, p = .431. In study 2, we utilized an additional test to check for non-verbal intelligence (i.e., Standard Progressive Matrices, SPM; Raven and De Lemos, 1958).

EEG measurements

Polysomnography (PSG) was recorded with SynAmps EEG amplifiers (Model 5083; NeuroScan Inc., El Paso, TX). Signal was filtered between 0.05 and 100 Hz and digitized with a 16-bit precision (0.084 μV/LSB accuracy, where LSB is least significant bit). Twenty-one Ag/AgCl electrodes (Fp1, Fpz, Fp2, F7, F3, Fz, F4, F8, T3, C3, Cz, C4, T4, T5, P3, Pz, P4, T6, O1, Oz, and O2) were placed on the scalp according to the international 10–20 system and were online referenced to FCz. In addition, we recorded two mastoid electrodes for later re-referencing, two vertical and two horizontal electrooculograms (EOGs), one bipolar sub-mental electromyogram (EMG), one bipolar electrocardiogram (ECG), and one bipolar respiratory channel. During the screening night, we used a reduced setup consisting of eight electroencephalograms (EEGs), four EOGs, and one bipolar ECG, but three bipolar EMGs (musculus mentalis, musculus tibialis left and right) and three bipolar respiratory channels (nasal airflow, chest, and abdominal). In study 2, we additionally utilized pulse oximetry on the index finger to better identify possible sleep disorders that are not related to insomnia. All signals were online digitized with 500-Hz sampling rate and offline filtered between 0.1 and 70 Hz, with an exclusion of 50-Hz notch frequency. Every night was fixed to 8-hr time in bed and later semi-automatically scored for sleep stages and SS by The Siesta Group© (Somnolyzer 24 × 7; Anderer et al., 2005; Anderer et al., 2004), according to the criteria of American Academy of Sleep Medicine (Iber, Ancoli-Israel, Chesson, and Quan, 2007).

Sleep spindles

In accordance with common practice, SS were detected based on the signal from the C3 electrode during non-rapid eye movement sleep (NREM, N2–N3 sleep stages). Advantages of analyzing signal from a central site are discussed elsewhere (Schabus et al., 2006). SS were detected with regards to the following criteria: (a) frequency between 11 and 15 Hz, (b) amplitude higher than 25 µV, (c) duration longer than 0.5 s, and (d) controlled for muscle (30–40 Hz) and alpha (8–12 Hz) artifacts (for details, refer to Anderer et al., 2005). Furthermore, we divided spindles into two groups: a slow range from 11 to 13 Hz and a fast range from 13 to 15 Hz. For all identified SS, we calculated two characteristics: (a) sleep spindle density (SpD) and (b) sleep spindle activity (SpA). SpD reflects the average number of SS detected in 1 min of sleep, whereas SpA reflects the average duration as well as amplitude of a spindle in a given participant (SpA = mean spindle duration × mean spindle amplitude) (see also Schabus et al., 2004).

Slow oscillations

SO were detected based on the signal from the Fz electrode during NREM. They were detected according to standard criteria (Massimini, Huber, Ferrarelli, Hill, and Tononi, 2004) with in house-built MATLAB routines (MathWorks®, Natick, MA) and fulfilling the following three criteria: (a) a negative zero crossing preceding a positive zero crossing in a time window between 0.3 and 1 s, (b) the peak negativity between those two zero crossings exceeding −80 µV, and (c) the peak-to-peak amplitude (between the negative and subsequent positive peaks) being higher than 140 µV. For all identified SO, we calculated two characteristics: (a) SO density (number of detections per 1 min of sleep) and (b) mean peak-to-peak amplitude (the amplitude from peak negativity – the strongest hyperpolarization – to peak positivity – the strongest depolarization).

Experimental design

Pre-screening was conducted using various questionnaires and phone consultations. If inclusion criteria were met, subjects reported to the laboratory for the study intake and spent their first (screening night) in the laboratory, which allowed them to adapt to the novel environment, and screen for sleep disorders that are not related to primary insomnia. Each experimental night was preceded by an encoding session – word pairs presented in randomized order on a computer screen. The encoding session started on average in study 1 around 10:00 p.m. and in study 2 around 9:45 p.m. A first cued recall session was performed in the evening after learning, prior to sleep. A second cued recall session was performed on the subsequent morning, 30 min after awakening. No caffeine intake was allowed during the entire day, on which a testing session took place. Each night, the PSG recordings started between 11 p.m. and midnight and continued for consecutive 8 hr (Fig.1).

Fig. 1.
Fig. 1.

Comparison of study 1 and study 2 designs. (a) Study 1 and (b) study 2. While study 1 was designed to investigate overnight memory change only, study 2 additionally focused on the susceptibility of memories to an interference manipulation in the morning. In addition, the two studies differed during evening encoding: while study 1 consisted of two encoding sessions (80 word pairs), study 2 required that participants learned to a 70% threshold. B1–60 or C21–60 refers to pairing of the 60 (or 40) words with words from a list A in study 2 (e.g., house–mouse [A21–B21] and then house–table [A21–C21])

Citation: Sleep Spindles & Cortical Up States Sleep Spindles & Cortical Up States 1, 1; 10.1556/2053.1.2016.001

In both studies, a screening night was followed by an experimental night. In study 1, the encoding session consisted of two presentations of 80 (A–B) word pairs. After a 10-min break, the subjects underwent the first cued retrieval session (evening retrieval), and in the morning, the second (morning retrieval) cued retrieval session of all A–B word pairs, without any feedback on correctness. In study 2, during the encoding session, participants studied 60 word pairs (A1–60–B1–60) and straight afterward underwent retrieval sessions with feedback that lasted until they reached a criterion of 70% (42 pairs) correctly recalled word pairs. Regardless of the 70% threshold, all 60 cues were presented at least once. If the participant did not reach the criterion in the first run, all incorrectly retrieved words were presented again until at least 70% of the 60 words were recalled correctly at the end of the recall block (i.e., theoretically evening recall can vary between 70% and 100%). At a subsequent morning retrieval of 20 (randomly chosen), A–B associations were tested with only the first (A) word being presented as cue (morning retrieval: A1–20–B1–20). Afterward, an interference session was introduced, where the remaining 40 words from list A were paired with 40 new words from a list C (A21–60–C21–60). Therefore, participants studied 40 new associations (A–C) which were already previously paired with an A word (e.g., original A–B pair: house–river; interference A–C pair: house–table). Training was analogous to the evening session (learning until at least 70% of the 40 A–C word associations were correctly retrieved). After a 20-min break, another cued retrieval session of the 40 A–B and A–C pairings was performed (post-interference retrieval: A21–60–B21–60 and A21–60–C21–60). Subjects were shown the A words as a cue and were asked to recall, if possible, both B and C words previously paired with the A cue word. After a delay of 5–8 days, a behavioral follow-up of the 60 A–B and 40 A–C associations was conducted (delayed retrieval: A1–60–B1–60 and A21–60–C21–60). All 24 patients and 24 out of 29 control subjects reported to the sleep laboratory for the delayed retrieval session.

Word-pair task

During the encoding session (Fig.2), word pairs were presented visually on a 15.6-in. computer screen. The words from a pre-existing list were arbitrarily paired by a computer algorithm and could at random be semantically related. For better control of the used mnemonic strategy, subjects were instructed to imagine a visual relation between two words of a pair. The differences between the two studies are also highlighted in the Supplementary Material (Table S1).

Fig. 2.
Fig. 2.

Comparison of the memory task in study 1 and study 2. (a) Study 1 and (b) study 2. The main difference between the two studies is a feedback about correctness during retrieval in study 2. Learning was terminated after participants reached at least 70% correct recalls (in study 2). Study 1 used 80 word pairs, whereas study 2 used 60 word pairs during the evening session

Citation: Sleep Spindles & Cortical Up States Sleep Spindles & Cortical Up States 1, 1; 10.1556/2053.1.2016.001

In study 1, during the encoding session, each of the 80 (A–B) word pairs was shown on a computer screen for 1,500 ms followed by 5,000 ms of a white centered fixation cross flipping to gray color for another 3,500 ms (Fig.2a). Whereas during the time of the white cross subjects should focus on encoding of the preceding word pair, the gray cross marked a period of mental relaxation. Encoding session of 80 word pairs was performed twice in a row (i.e., encoding 1 and encoding 2; Fig.1a). In total, the encoding sessions lasted around 29 min and were followed by a 10-min break. During the retrieval sessions, the first (A) word from each pair (cue) was presented on a screen until participant pressed a button and answered orally – with the second (B) word from the pair, or for a maximum delay of 6,500 ms (Fig.2a).

In study 2, during encoding, each of the 60 (A1–60–B1–60) word pairs was shown on a computer screen for 7 s. Immediately after learning, subjects performed a cued recall session, where the first word (A) from each pair (cue) was presented on a screen until participant pressed a button and answered orally – with the second (B) word, or for a maximum delay of 10 s. In the latter case, as well as in the case of incorrect responses, the correct answer (in form of the whole word pair) was displayed on the screen for 2 s (Fig.2b). For study task details also see the Supplementary Material.

Baseline memory performance

In this paper, we primarily focused on the effects of BMP, that is, the efficiency of declarative learning in the evening on overnight consolidation. In study 1, BMP is calculated as the percentage of successfully recalled (A–B) word pairs during the evening retrieval. Similarly, in study 2, BMP is calculated as the percentage of successfully recalled (A–B) word pairs after the first (evening) retrieval cycle. For all subsequent analyses, each study sample was separated into two BMP groups by a simple median split. We included two insomnia samples (study 1 and study 2) as well as one healthy control sample (from study 2) in our analyses.

Statistical analysis

Statistical analyses were performed using SPSS 18 software (SPSS Inc., Chicago, IL). As the main analysis was exploratory in nature, we conducted two-tailed level of significance tests at a set of p-value <.05. When assumptions for parametric testing were violated, non-parametric tests were utilized.

For better characterizing the BMP measure, we not only test associations of OMC (= morning retrieval score − evening retrieval score) with BMP but also its association with age as well as general cognitive ability measures (i.e., study 1: WMS-R score, study 2: SPM score). We do this by using correlation analyses across the study participants and for both studies separately (i.e., study 1: N = 23, study 2: N = 52).

To better understand the contribution of variables in explaining variance in OMC, we performed forward stepwise linear regression analysis. In the analysis, we included all three study samples (N = 76) and the following predictors: age, gender, BMP, sum of percentage of N2 and N3 sleep (percentage of N2 and N3 had large zero-order correlation: r = .62, p < .001), density and intensity of slow as well as fast SS, and amplitude and density of SO.

To investigate the influence of BMP on OMC, we utilized a mixed-design two-way analysis of variance (ANOVA) with between-group factor PERFORMANCE (BMP+, BMP−), the repeated measure OMC-RETRIEVAL (evening, morning), and the dependent variable RETRIEVAL SCORE (in %) in study 1. For study 2, we performed a three-way ANOVA with an additional between-group factor SAMPLE (healthy, insomnia). To explore significant effects, we utilized post-hoc paired-sample t-tests comparing retrieval scores from evening and morning for the BMP groups separately [adjusted with the False Discovery Rate (FDR) procedure]. We additionally performed a control analysis of BMP and OMC association by dividing (median split of OMC measure) each study sample separately into OMC+ and OMC− subgroups. We then compared with independent-sample t-tests BMP values of two subgroups from each study sample.

For analyzing our interference manipulations (interference by overwriting and interference of time delay) in study 2, we conducted a three-way ANOVA with the two between-group factors PERFORMANCE (BMP+, BMP−) as well as SAMPLE (healthy, insomnia), and the repeated measure INTERFERENCE-RETRIEVAL with dependent measure RETRIEVAL SCORE CHANGE (in %). The repeated measure had two levels: interference change (= post-interference retrieval score − morning retrieval score) and delay change (= delayed retrieval score − post-interference retrieval score). For further investigation of observed effects, we conducted independent-sample t-tests comparing interference change and delay change between the two BMP groups, for the two samples separately. In the last step, we tracked changes over consecutive retrievals, in each BMP group separately, with paired-sample t-tests.

To investigate the differences in SS measures in study 1, we performed 2 two-way repeated measure ANOVAs with a group variable PERFORMANCE (BMP+, BMP−), repeated measure SPINDLE TYPE (fast, slow), and the dependent variable SpA or SpD. Differences in SO measures were investigated in study 1 with two-way repeated measure ANOVAs with a group variable PERFORMANCE and repeated measure OSCILLATION MEASURE (amplitude, density). In study 2, we investigated the differences in SS and SO with three-way ANOVAs with an additional group factor SAMPLE (healthy, insomnia). Post-hoc independent-sample t-tests were used to further explore the differences between BMP groups (FDR corrected). The effects of SAMPLE in study 2 were addressed with independent-sample t-tests (insomnia vs. healthy).

Differences between BMP groups in the sleep architecture were investigated with three multivariate ANOVAs (MANOVAs) using Pillai’s Trace, with the between-group factor PERFORMANCE (BMP+, BMP−) and percentage of three sleep stages [N2, N3, and rapid eye movement sleep (REM)] as the dependent variables. Follow-up analyses were performed with one-way ANOVA on each dependent variable.

To disentangle effects of BMP from the effect of age, we further investigated the statistically significant results using (a) partial correlations (where effects of age were held constant) and (b) mediation analysis. In the mediation analysis, we used BMP as the independent variable and age as the mediator variable, and compared total effects with direct effects (i.e., the contribution of BMP directly mediating overnight memory consolidation).

Results

We analyzed data from three study samples: (a) insomnia sample from study 1, (b) insomnia sample from study 2, and (c) healthy sample from study 2. All participants performed word-pair (A–B) declarative memory task in the evening (evening retrieval) before as well as in the morning (morning retrieval) after 8 hr of nocturnal sleep. Based on their performance during the evening retrieval session, we calculated their individual trait measure of BMP. Each of the three study samples was divided with median split into two BMP groups: moderate (BMP−) and high (BMP+) performers. It is to note that the design of the two studies included in our analyses slightly differed. Especially, while all the participants from the study 1 underwent in the evening the same amount of learning and retrieval session, participants from the study 2 were retrieving word pairs until they reach the criterion of 70% correctly recalled material. Moreover, in the two studies, participants were requested to encode different amounts of the word pairs: 80 in the study 1 and 60 in the study 2. In this paper, we investigated differences in memory retrieval in BMP groups with respect to changes in sleep architecture, SpA, and SpD as well as density and peak-to-peak amplitude of SO.

Demographics

Demographic information of our two studies and six BMP groups is depicted in Table1. Comparing the two insomnia samples (from study 1 vs. from study 2), for each BMP category separately, we found that none of the groups differed significantly with respect to age or the PSQI score. However, data indicate a difference in sleep efficiency (SE) between the two BMP− groups of the insomnia subgroups (t21 = 2.380, p = .027), with participants from study 1 having more efficient sleep.

Table 1.

Demographic information across studiesa

article image

Note. PSQI, Pittsburgh Sleep Quality Index Questionnaire; BMP, baseline memory performance; SE, sleep efficiency; OMC, overnight memory change.

In all three study samples, the two BMP groups differed significantly in OMC with BMP− groups exhibiting stronger forgetting overnight. In both insomnia samples, the two BMP− groups have higher mean age as well as mean PSQI score. Yet note the difference in OMC even in healthy individuals (study 2) with identical age in both BMP− and BMP+ groups. M ± SD is reported.

Baseline memory performance

Exploratory linear regression analysis performed over all three study samples (N = 76) revealed that only two predictors significantly contributed into explaining variance in OMC: BMP (t = 5.44, p < .001) and slow SpA (t = 2.06, p = .043). This model significantly explained variance in OMC (F2,75 = 15.24, p < .001, adjusted R2 = .275). Age as a predictor was excluded from this model (t = 1.23, p = .222). Furthermore, we compared associations of OMC with BMP and age, and as BMP was also moderately correlated with age (study 1: r = −.40, p = .060, study 2: rs = −.34, p = .013), we checked for the relative contribution of these two variables in explaining OMC in the two studies separately. BMP correlated robustly with OMC in study 1 (r = .63, p = .001) as well as in study 2 (r = .55, p < .001) (explaining 39% and 30% of variance in OMC in the two studies, respectively). Age itself revealed weaker relations to OMC (study 1: r = −.44, p = .036, study 2: r = −.27, p = .052) and explained less variance in OMC (study 1: 19%, study 2: 7%). Yet the correlation between OMC and BMP was statistically significant and basically unchanged when controlling for age (study 1: r = .55, p = .008; study 2: r = .50, p < .001).

Also our control analysis revealed that in all three study samples, OMC+ and OMC− subgroups differed significantly in BMP measure in study 1 (t21 = 3.30, p = .003, OMC+ group: BMP = 76.9 ± 15.0, OMC = 2.5 ± 2.7; OMC− group: BMP = 48.1 ± 24.2, OMC = −4.4 ± 2.7), insomnia sample from study 2 (t22 = 2.28, p = .033, OMC+ group: BMP = 64.9 ± 17.3, OMC = 8.3 ± 6.5; OMC− group: BMP = 49.6 ± 15.3, OMC = −11.4 ± 9.9) as well as healthy sample from study 2 (t27 = 2.52, p = .018, OMC+ group: BMP = 59.9 ± 20.6, OMC = 4.6 ± 6.4; OMC− group: BMP = 41.6 ± 18.5, OMC = −21.0 ± 11.6).

Consequently, we are concentrating on BMP rather than age as grouping variable in all of the following analyses. For better understanding the nature of the utilized BMP measure, we also correlated it with trait scores of memory aptitude and intelligence of study 1 and study 2, respectively. As expected, the results indicated a correlation of evening memory recall and a general test battery of memory aptitude (study 1: WMS-R) (WMS-R score; r = .37, p = .036). General cognitive ability or intelligence (study 2: SPM) on the other hand showed a smaller trend-like association (SPM raw score; r = .23, p = .095).

Behavioral overnight memory change

Participants in study 1 and study 2 started with different amounts of to be encoded material in the evening (study 1: 80 word pairs, study 2: 60 word pairs). Given that, insomnia patients from study 1 encoded the highest number of word pairs in the evening if they belonged to the BMP+ group (M = 66.5, SD = 8.4), whereas participants from the BMP− group (study 1) encoded the lowest absolute number of word pairs (M = 32.5, SD = 10.9) as compared with any group of study 2. Due to the 70% (evening) learning criterion in study 2, insomnia patients (BMP+: M = 48.3, SD = 4.2; BMP−: M = 48.0, SD = 3.3) as well as healthy controls (BMP+: M = 50.5, SD = 4.6; BMP−: M = 48.1, SD = 2.5) went to bed with comparable numbers of encoded word pairs.

In general, our behavior analyses found a consistent pattern across all three study samples, that is, BMP+ groups did not reveal any significant forgetting from evening to morning, whereas BMP− tended to forget overnight in all three study samples (Fig.3). Thus, we found OMC-RETRIEVAL × PERFORMANCE interactions to be significant in study 1 (F1,21 = 14.993, p = .001, ηp2=.417) as well as in study 2 (F1,49 = 12.858, p = .001, ηp2=.204). Post-hoc FDR-corrected tests revealed no significant forgetting (OMC) in any of the BMP+ groups (insomnia sample from study 1: t10 = 1.507, p = .249; insomnia sample from study 2: t12 = 1.476, p = .249; healthy sample from study 2: t14 = .269, p = .792). However, forgetting was evident in BMP− groups of the insomnia sample from study 1 (t11 = 3.930, p = .007) and the healthy sample from study 2 (t13 = 3.863, p = .007). In the insomnia sample from study 2, the overnight forgetting was not significant (t10 = 1.308, p = .264).

Fig. 3.
Fig. 3.

Interaction of overnight memory change and BMP. (a) Study 1 – insomnia sample. The BMP+ group generally retrieved more word pairs during both recall sessions (evening + morning) than the BMP− group, with only the BMP− group exhibiting overnight forgetting. (b) Study 2 – insomnia sample. BMP+ and BMP− groups revealed an insignificant pattern with only the BMP− group forgetting overnight. (c) Study 2 – healthy sample. Similarly like in the insomnia sample from study 1, only the BMP− group significantly forgot overnight. Values are always expressed as relative percentage, separately for the two BMP groups and three study samples. Each bar depicts the mean value ± the standard error. In study 2, the blue bars (score for evening retrieval) depict the mean values of the final retrieval score after the last learning cycle (therefore exceeding the 70% threshold). BMP, baseline memory performance

Citation: Sleep Spindles & Cortical Up States Sleep Spindles & Cortical Up States 1, 1; 10.1556/2053.1.2016.001

Controlling for age, a partial correlation between OMC and BMP was significant in the insomnia sample from study 1 (r = .550, p = .008) and in the healthy sample from study 2 (r = .525, p = .004), and revealed a trend toward significance in the insomnia sample from study 2 (r = .407, p = .054) suggesting a negligible influence of age on the BMP to OMC association.

A mediation analysis confirms that the direct effect of BMP on OMC (study 1 insomnia sample: b = .095, t = 2.94, p = .008; study 2 insomnia sample: b = .394, t = 2.04, p = .054; study 2 healthy sample: b = .386, t = 3.14, p = .004) is comparable to the total effect, which includes age and BMP (study 1 insomnia sample: b = .110, t = 3.70, p = .001; study 2 insomnia sample: b = .358, t = 2.69, p =.013; study 2 healthy sample: b = .412, t = 3.14, p = .004). (Note that the total effect is identical to the simple correlations, in this case the correlation between BMP and OMC.)

An additional analysis focusing on the susceptibility to interference effects (study 2) revealed an association of BMP with morning to post-interference retrieval change (r = .451, p = .027), only when age was not controlled (r = .264, p = .175) (for more details, see the Supplementary Material).

Sleep parameters and sleep quality

We checked for the differences in sleep architecture by calculating MANOVAs with the between-group factor PERFORMANCE and percentage of three sleep stages (N2, N3, and REM) as the dependent variables. We found Pillai’s Trace statistic to be significant only in the study 1 (V = .48, F3,19 = 5.792, p = .005, ηp2=.478). Separate univariate ANOVAs on the outcome variables revealed that the difference between the two BMP groups from the study 1 was significant only in the relative amount of N3 sleep (F1,21 = 16.746, p = .001, ηp2=.444), with BMP+ having more N3% (Fig.4). Also SE did not differ between the two BMP groups from the study 1 insomnia sample (t21 = 0.398, p = .694). In study 2, however, BMP+ groups tended to have higher SE (insomnia sample: t22 = 3.549, p = .004; healthy sample: t27 = 2.006, p = .055) (Table1).

Fig. 4.
Fig. 4.

Sleep architecture. Plots depict the percentage of TIB for each of the six groups and every sleep stage. The only significant difference revealed was found in the study 1 insomnia sample, that is, the BMP− group had significantly less N3% than the BMP+ group. TIB, time in bed; R, rapid eye movement sleep; BMP, baseline memory performance

Citation: Sleep Spindles & Cortical Up States Sleep Spindles & Cortical Up States 1, 1; 10.1556/2053.1.2016.001

A partial correlation controlling for age revealed a significant association between BMP and the relative amount of N3 sleep in the insomnia sample from study 1: r = .499, p = .018. A mediation analysis confirmed this significant relationship with both total (b = .047, t = 3.46, p = .002) and direct (b = .035, t = 2.58, p = .018) effects (i.e., direct effect of BMP in the evening on N3 sleep thereafter) being significant.

Sleep spindles and slow oscillations

Analysis of SpD revealed an overall pattern of higher density of fast, but not slow SS in BMP+ as compared with BMP−, which was especially pronounced in study 2 (Fig.5a). In study 1, the PERMORMANCE × SPINDLE TYPE interaction for SpD turned out to be not significant (F1,21 = .127, p = .725), while in study 2 it reached significance (F1,49 = 8.645, p = .005, ηp2=.150). Post-hoc FDR-corrected tests revealed a trend toward higher fast SpD in the BMP+ than in the BMP− group in the healthy sample of study 2 (t27 = 2.516, p = .054). In contrast to SpD, analysis of SpA revealed a general pattern of higher activity of both SS types in BMP+ groups. Yet, the PERFORMANCE × SPINDLE TYPE interaction for SpA was not significant in any of the three study samples (Fig.5b). The main effect for factor PERFORMANCE was significant in the study 1 insomnia sample (F1,21 = 4.946, p = .037, ηp2=.191) as well as in the study 2 healthy sample (F1,27 = 5.929, p = .022, ηp2=.180) suggesting general higher SpA in BMP+ groups. In the study 2 insomnia sample, we did not observe a significant main effect for PERFORMANCE: F1,22 = 1.581, p = .222, ηp2=.067. Post-hoc FDR-corrected tests revealed that the BMP+ group from the study 2 healthy sample tended to have significantly higher slow (U = 59.0, z = 2.008, p = .075) as well as fast (U = 54.0, z = 2.226, p = .078) SpA. Similarly, the BMP+ group from the study 1 insomnia sample revealed trends toward higher slow SpA (t21 = 2.082, p = .074). Only in the insomnia sample of study 2 no significant differences between the two BMP groups were observed.

Fig. 5.
Fig. 5.

Sleep spindle density and activity and its relation to BMP. (a) depicts slow and fast sleep spindle density (SpD) and (b) depicts slow and fast sleep spindle activity (SpA) for both BMP groups and across both studies. Results reveal an overall pattern of higher fast SpD as well as general higher SpA in individuals with high than low BMP. Bars depict mean ± standard error. SpD, sleep spindle density; SpA, sleep spindle activity; BMP, baseline memory performance

Citation: Sleep Spindles & Cortical Up States Sleep Spindles & Cortical Up States 1, 1; 10.1556/2053.1.2016.001

Bivariate correlations between BMP and SS were significant only in the control sample from study 2 and only for fast spindles (density: r = .414, p = .026; activity: r = .393, p = .035). The associations remained (marginally) significant when controlling for age (density: r = .377, p = .048; activity: r = .351, p = .067). Similarly, a mediation analysis revealed modest differences in total (density: b = .031, t = 2.36, p = .026; activity: b = .061, t = 2.22, p = .035) and direct (density: b = .027, t = 2.07, p = .048; activity: b = .052, t = 1.91, p = .067) effects.

Finally, we analyzed the differences in the SO features. Analysis of SO density and peak-to-peak amplitude, however, revealed no association with BMP after controlling for age.

Effects of insomnia – study 2

In the last step, we compared the healthy and insomnia samples from study 2, to investigate the effects of decreased subjective sleep quality. Across the whole sample (insomnia: N = 24, healthy: N = 29), we found that patients compared with healthy controls had significantly higher PSQI score (U = 0.0, z = 6.251, p < .001) and by trend lower SE (U = 248.5, z = 1.778, p = .075), but the two samples did not differ significantly in age or BMP. With respect to behavioral overnight memory change, we did not observe any significant difference between healthy and insomnia samples (no significant OMC-RETRIEVAL × PERFORMANCE × SAMPLE interaction, nor PERFORMANCE × SAMPLE interaction, nor a main effect for SAMPLE). Neither did the analysis of interference and time delay manipulation reveal any significant differences in memory performance between the two samples (no significant PERFORMANCE × SAMPLE × INTERFERENCE-RETRIEVAL interaction nor SAMPLE × INTERFERENCE-RETRIEVAL interaction). However, a main effect for SAMPLE (F1,44 = 7.747, p = .008, ηp2=.150) indicated that the healthy sample forgets on average less across the interference manipulation (−10.9%) than the insomnia sample (−16.2%). In addition to that we did not find any differences between the two samples (spindle and SO measures).

Discussion

In this paper, we intended to test the influence of a general cognitive trait reflecting good versus bad declarative learning (evaluated as encoding efficiency of word pairs during initial memory formation; termed “baseline memory performance”; BMP) on overnight memory consolidation in a large sample of healthy individuals (N = 29) and insomnia patients (N = 47). The goal was to identify the influence of this memory trait on two variants of a declarative word-pair association task and identify its potential relation to sleep features.

In general, data indicate that the general BMP trait is crucially related to overnight memory consolidation. Furthermore, we reveal higher spindle measures in good BMP performers and do find these relationships rather independent of the exact study design, sleep complaint, or age.

Differences between BMP groups

More specifically, our main behavioral finding is a significant influence of BMP on overnight forgetting irrespective of subjective sleep complaints. We found significant overnight forgetting of BMP− subjects in two out of three samples with effects of the third sample going in similar direction. BMP+ subjects showed overnight stabilization or consolidation of performance. Similarly, Tucker and Fishbein (2008) found sleep-dependent improvement (across a 45-min nap) in only those participants that already had higher performance during an initial training session. It has been proposed that individuals with higher cognitive abilities or intelligence are able to create stronger “offline” associations between learnt word pairs (Fenn and Hambrick, 2011; Schabus et al., 2008; Tucker and Fishbein, 2009) as also indicated by greater hippocampal activation at encoding. Specifically, this stronger hippocampal activation at encoding has been linked to stronger reactivations during subsequent sleep and as a consequence stronger sleep-dependent memory consolidation (Peigneux et al., 2004; Rauchs et al., 2011). Yet earlier studies from our own group failed to reveal a clear association of overnight declarative memory change and general tests of cognitive (Advanced Progressive Matrices – APM) or memory ability (WMS-R) (Heib et al., 2013).

In two out of three study samples, we observed significantly higher spindle activity, in BMP+ compared with BMP− groups. Higher density of (especially fast) spindles in BMP+ group was observed only in the healthy sample. As in this paper only BMP− groups seemed to forget overnight, our spindle results are in line with earlier findings linking SS to OMC of declarative memories. Specifically, earlier studies have revealed positive associations of SpA enhancement (from baseline to post-learning night) and OMC (Heib et al., 2015; Schabus et al., 2004). The recent results of Heib et al. (2015) suggested this association to be mediated by event-related theta power, already observable during pre-sleep memory recall. In addition, SpD has been shown to increase during sleep after learning but not after a non-learning control task and that better learners in general have higher SpD (Gais, Helms, Wagner, Mölle, and Born, 2002). Furthermore, higher SpA has been found in participants with higher cognitive abilities as measured using the APM test of intelligence or the WMS-R test of memory aptitude (Schabus et al., 2006; Schabus et al., 2008). As a consequence of these findings, (especially fast) SS have been proposed as a by-product of effective thalamocortical and intracortical connectivity, which seems to be the base for higher cognitive ability and supports more efficient memory encoding (Bodizs et al., 2005; Clemens, Fabo, and Halasz, 2005; Schabus et al., 2006).

It is therefore likely that in our study, the BMP+ groups established stronger neuronal links for the encoded word pairs, and therefore also consolidated these memories more efficiently overnight. It is important to note that a formerly raised issue by Stickgold (2004) – namely that presumably only individuals with higher baseline SpA further enhance spindles and consolidate successfully overnight – might therefore be partly warranted but can only be revealed when a specific trait measure of word-pair encoding efficiency (such as the used BMP measure) is taken into account rather than general measures of cognitive ability (such as APM or WMS-R).

Differences between the two BMP groups in SpA were statistically significant only for the healthy sample from study 2 and the insomnia sample from study 1. Yet interestingly, the BMP− group that did not forget significantly overnight (study 2 insomnia) also not showed a decreased SpA (as compared with BMP+). This observation supports the idea of the functional role of SS for overnight memory consolidation; and suggests that BMP+ may experience more pronounced reactivation processes during sleep which might stem from a more efficient encoding prior to sleep. As a consequence, new memories in BMP+ get stabilized overnight and hence accessible for the long term. Furthermore, the only difference in sleep architecture between the two BMP groups was found in the relative amount of N, with BMP+ subjects of study 1 showing more deep slow wave sleep post-learning. It is interesting to note that we also found the largest difference in behavior exactly between these two BMP groups. Therefore, we speculate that the increased relative amount of N3 sleep is reflecting the higher “initial memory load” in that group. Note that although all three BMP+ groups of study 1 and study 2 retrieved comparable relative amounts of information in the evening (around 80%), this amounted to about 66 word pairs in study 1 as compared with around 49 word pairs in study 2. We speculate that these subtle differences of demand on memory, that is, to be remembered information just before falling asleep, may determine whether sleep architecture changes become evident in the data. This interpretation is also in line with the synaptic homeostasis hypothesis (Tononi and Cirelli, 2006), which would suggest that a higher pre-sleep memory load should result in a more intensive downscaling process manifested in a higher density of SO (or N3 sleep) during the post-learning night. It is to note that “initial memory load” did not affect any of the other before mentioned results (OMC or SS).

Effects of insomnia

With respect to insomnia, we observed surprisingly little differences in memory performance or sleep measures. Also note that earlier studies failed to find memory decrements in insomnia patients (for review, see Cipolli, Mazzetti, and Plazzi, 2013). This might be related to the effect that robust differences in sleep architecture or SO intensity are often absent in insomnia studies rendering it more a subjective rather than objectively founded complaint of unrestful sleep (Lichstein, Wilson, Noe, and Aguillard, 1994).

Limitations for the interpretation

Finally, it is to note that the BMP measure used here was only moderately correlated with general trait measures, such as WMS-R or SPM. Therefore, the utilized BMP measure used in our study does not evaluate general cognitive or memory abilities but rather assesses the specific efficiency in learning declarative word-pair associations. Furthermore, we here analyzed participants with a wider age range than in comparable studies which might influence the results, as OMC (Mander et al., 2013), SS (Crowley, Trinder, Kim, Carrington, and Colrain, 2002), and SO (Fogel et al., 2012) are well known to change with progressive age. Yet, as outlined earlier, it appears that BMP rather than age is the crucial driving force when trying to understand sleep-dependent memory consolidation as confirmed by mediation analysis and partial correlations. Also note that the revealed effect of significantly lower overnight forgetting in BMP+ groups is observed in that study sample, where the two BMP groups are of identical age, that is, in the healthy sample from study 2.

It should be also noted that our study samples were dominated by females. We have not observed the effects of gender in our findings, although we cannot exclude that the results might be influenced by the variability in menstrual cycle, menopause, or oral contraceptives.

Many of the discussed effects become statistically significant only in some subgroups, yet with all groups pointing in the same direction. Importantly, if subjects are pooled together, we verify these findings statistically and therefore dared to interpret them as general findings.

Summary

Taken together, our results revealed (declarative) overnight forgetting exclusively in participants who also showed a lower capacity to learn such word-pair associations prior to sleep in the first place. In addition, sleep features such as SS appear to be generally higher in good compared with less efficient learners, independent of study design, or insomnia complaints.

Authors’ contribution

MW and MS wrote the paper. MW, DPJH, HG, and MS analyzed data. DPJH contributed analysis tools. DPJH, HG, KH, MW, and MS preformed research. MS and KH designed and supervised the research.

Conflict of interest

The authors declare that the research was conducted in the absence of any commercial or financial relationship that could be construed as a potential conflict of interest.

Ethics

This study was conducted in accordance with the ethical principles in the Declaration of Helsinki and was approved by the local research ethics committee (University of Salzburg). Before participation, all subjects gave written informed consent.

Acknowledgements

This study was financially supported by the FWF project P-21154-B18. D.P.J. Heib and M. Wislowska were supported by the Doctoral College “Imaging the Mind” (FWF; W1233-G17) as well as I-934-B23 (D.P.J. Heib). The authors would like to thank Wiebke Boening, Katharina Engl, Martina Feichtinger, Cornelia von Gamm, Nikolina Luketina, Tina Moeckel, Marit Petzka, Daniela Tschann, Gabriela Werner, and Micheala Wittek for their support in this study.

References

  • Anderer P , Gruber G , Parapatics S , Woertz M , Miazhynskaia T , Klösch G , Saletu B , Zeitlhofer J , Barbanoj MJ , Danker-Hopfe H , Himanen S-L , Kemp B , Penzel T , Grözinger M , Kunz D , Rappelsberger P , Schlögl A , and Dorffner G (2005). An E-health solution for automatic sleep classification according to Rechtschaffen and Kales: validation study of the Somnolyzer 24 × 7 utilizing the Siesta database. Neuropsychobiology 51:115133.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Anderer P , Saletu B , Saletu-Zyhlarz GM , Gruber G , Parapatics S , Miazhynskaia T , Woertz M , Klösch G , Zeitlhofer J , and Dorffner G (2004). Recent advances in the electrophysiological evaluation of sleep. In W Drinkenburg, G Ruigt, and M Jobert (Eds.), Essentials and applications of EEG research in preclinical and clinical pharmacology, (pp 307339). Berlin: Verlag für Studium & Praxis OHG.

    • Search Google Scholar
    • Export Citation
  • Beck AT , Epstein N , Brown GK , and Steer RA (1988). An inventory for measuring clinical anxiety: psychometric properties. J Consult Clin Psychol 56:893897. doi:10.1037/0022-006X.56.6.893

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Beck AT , Steer RA , and Brown GK (1996). Manual for the beck depression inventory-II. San Antonio, TX: Psychological Corporation.

  • Bodizs R , Kis T , Lazar AS , Havran L , Rigo P , Clemens Z , and Halasz P (2005). Prediction of general mental ability based on neural oscillation measures of sleep. J Sleep Res 14:285292. doi:10.1111/j.1365-2869.2005.00472.x

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Buysse DJ , Reynolds CF , Monk TH , Berman SR , and Kupfer DJ (1989). The Pittsburgh Sleep Quality Index: a new instrument for psychiatric practice and research. Psychiatry Res 28:193213. doi:10.1016/0165-1781(89)90047-4

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Cipolli C , Mazzetti M , and Plazzi G (2013). Sleep-dependent memory consolidation in patients with sleep disorders. Sleep Med Rev 17:91103. doi:10.1016/j.smrv.2012.01.004

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Clemens Z , Fabo D , and Halasz P (2005). Overnight verbal memory retention correlates with the number of sleep spindles. Neuroscience 132:529535. doi:10.1016/j.neuroscience.2005.01.011

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Clemens Z , Molle M , Eross L , Barsi P , Halasz P , and Born J (2007). Temporal coupling of parahippocampal ripples, sleep spindles and slow oscillations in humans. Brain 130:28682878. doi:10.1093/brain/awm146

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Crowley K , Trinder J , Kim Y , Carrington M , and Colrain IM (2002). The effects of normal aging on sleep spindle and K-complex production. Clin Neurophysiol 113:16151622. doi:10.1016/S1388-2457(02)00237-7

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Diekelmann S and Born J (2010). The memory function of sleep. Nat Rev Neurosci 11:114126. doi:10.1038/nrn2762

  • Edinger JD , Bonnet MH , Bootzin RR , Doghramji K , Dorsey CM , Espie CA , Jamieson AO , McCall WV , Morin CM , and Stepanski EJ (2004). Derivation of research diagnostic criteria for insomnia: report of an American Academy of Sleep Medicine Work Group. Sleep 27:15671596.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Ego-Stengel V and Wilson MA (2010). Disruption of ripple-associated hippocampal activity during rest impairs spatial learning in the rat. Hippocampus 20:110. doi:10.1002/hipo.20707

    • Search Google Scholar
    • Export Citation
  • Fenn KM and Hambrick DZ (2011). Individual differences in working memory capacity predict sleep-dependent memory consolidation. J Exp Psychol Gen 141:404410. doi:10.1037/a0025268

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Fogel SM , Martin N , Lafortune M , Barakat M , Debas K , Laventure S , Latreille V , Gagnon J-F , Doyon J , and Carrier J (2012). NREM sleep oscillations and brain plasticity in aging. Front Neurol 3:176. doi:10.3389/fneur.2012.00176

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Fogel SM and Smith CT (2011). The function of the sleep spindle: a physiological index of intelligence and a mechanism for sleep-dependent memory consolidation. Neurosci Biobehav Rev 35:11541165. doi:10.1016/j.neubiorev.2010.12.003

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Gais S , Helms K , Wagner U , Mölle M , and Born J (2002). Sleep spindle density increases after learning a declarative memory task. J Psychophysiol 16:237238.

    • Search Google Scholar
    • Export Citation
  • Girardeau G , Benchenane K , Wiener SI , Buzsáki G , and Zugaro MB (2009). Selective suppression of hippocampal ripples impairs spatial memory. Nat Neurosci 12:12221223. doi:10.1038/nn.2384

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Griessenberger H , Heib DPJ , Lechinger J , Luketina N , Petzka M , Moeckel T , Hoedlmoser K , and Schabus M (2013). Susceptibility to declarative memory interference is pronounced in primary insomnia. PLoS One 8:e57394. doi:10.1371/journal.pone.0057394

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Heib DPJ , Hoedlmoser K , Anderer P , Gruber G , Zeitlhofer J , and Schabus M (2015). Oscillatory theta activity during memory formation and its impact on overnight consolidation: a missing link? J Cogn Neurosci 27:16481658. doi:10.1162/jocn_a_00804

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Heib DPJ , Hoedlmoser K , Anderer P , Zeitlhofer J , Gruber G , Klimesch W , and Schabus M (2013). Slow oscillation amplitudes and up-state lengths relate to memory improvement. PLoS One 8:e82049. doi:10.1371/journal.pone.0082049

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Iber C , Ancoli-Israel S , Chesson AL , and Quan SF (2007). The AASM manual for the scoring of sleep and associated events: rules, terminology and technical specifications. Darien, IL: American Academy of Sleep Medicine.

    • Search Google Scholar
    • Export Citation
  • Lichstein KL , Wilson NM , Noe SL , and Aguillard R (1994). Daytime sleepiness in insomnia: behavioral, biological and subjective indices. Sleep 17:693702.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Mander BA , Rao V , Lu B , Saletin JM , Lindquist JR , Ancoli-Israel S , Jagust W , and Walker MP (2013). Prefrontal atrophy, disrupted NREM slow waves and impaired hippocampal-dependent memory in aging. Nat Neurosci 16:357364. doi:10.1038/nn.3324

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Massimini M , Huber R , Ferrarelli F , Hill S , and Tononi G (2004). The sleep slow oscillation as a traveling wave. J Neurosci 24:68626870. doi:10.1523/JNEUROSCI.1318-04.2004

    • Crossref
    • Search Google Scholar
    • Export Citation
  • McClelland JL , Mcnaughton BL , and O’Reilly RC (1995). Why there are complementary learning systems in the hippocampus and neocortex: insights from the successes and failures of connectionist models of learning and memory. Psychol Rev 102:419457. doi:10.1037/0033-295X.102.3.419

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Mölle M , Yeshenko O , Marshall L , Sara SJ , and Born J (2006). Hippocampal sharp wave-ripples linked to slow oscillations in rat slow-wave sleep. J Neurophysiol 96:6270. doi:10.1152/jn.00014.2006

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Nader RS and Smith CT (2001). The relationship between stage 2 sleep spindles and intelligence. Sleep 24:A160.

  • Peigneux P , Laureys S , Fuchs S , Collette F , Perrin F , Reggers J , Phillips C , Degueldre C , Fiore GD , Aerts J , Luxen A , and Maquet P (2004). Are spatial memories strengthened in the human hippocampus during slow wave sleep? Neuron 44:535545. doi:10.1016/j.neuron.2004.10.007

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Rauchs G , Feyers D , Landeau B , Bastin C , Luxen A , Maquet P , and Collette F (2011). Sleep contributes to the strengthening of some memories over others, depending on hippocampal activity at learning. J Neurosci 31:25632568. doi:10.1523/JNEUROSCI.3972-10.2011

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Raven JC and De Lemos MM (1958). Standard progressive matrices. London: Lewis.

  • Rosanova M and Ulrich D (2005). Pattern-specific associative long-term potentiation induced by a sleep spindle-related spike train. J Neurosci 25:93989405. doi:10.1523/JNEUROSCI.2149-05.2005

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Schabus M , Gruber G , Parapatics S , Sauter C , Klosch G , Anderer P , Klimesch W , Saletu B , and Zeitlhofer J (2004). Sleep spindles and their significance for declarative memory consolidation. Sleep 27:14791485.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Schabus M , Heib DPJ , Lechinger J , Griessenberger H , Klimesch W , Pawlizki A , Kunz AB , Sterman BM , and Hoedlmoser K (2013). Enhancing sleep quality and memory in insomnia using instrumental sensorimotor rhythm conditioning. Biol Psychol 95:126134. doi:10.1016/j.biopsycho.2013.02.020

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Schabus M , Hoedlmoser K , Gruber G , Sauter C , Anderer P , Klösch G , Parapatics S , Saletu B , Klimesch W , and Zeitlhofer J (2006). Sleep spindle-related activity in the human EEG and its relation to general cognitive and learning abilities. Eur J Neurosci 23:17381746. doi:10.1111/j.1460-9568.2006.04694.x

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Schabus M , Hoedlmoser K , Pecherstorfer T , Anderer P , Gruber G , Parapatics S , Sauter C , Kloesch G , Klimesch W , Saletu B , and Zeitlhofer J (2008). Interindividual sleep spindle differences and their relation to learning-related enhancements. Brain Res 1191:127135. doi:10.1016/j.brainres.2007.10.106

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Schramm E , Hohagen F , Grasshoff U , Riemann D , Hajak G , Weess HG , and Berger M (1993). Test–retest reliability and validity of the structured interview for sleep disorders according to DSM-III-R. Am J Psychiatry 150:867872. doi:10.1176/ajp.150.6.867

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Siapas AG and Wilson MA (1998). Coordinated interactions between hippocampal ripples and cortical spindles during slow-wave sleep. Neuron 21:11231128. doi:10.1016/S0896-6273(00)80629-7

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Spielberger CD (1983). Manual for the state-trait anxiety inventory (STAI). Palo Alto, CA: Consulting Psychologists Press.

  • Stickgold R (2004). Dissecting sleep-dependent learning and memory consolidation. Comment on Schabus M et al. Sleep spindles and their significance for declarative memory consolidation. Sleep 27:14431445.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • The WHOQOL Group (1998). Development of the World Health Organization WHOQOL-BREF quality of life assessment. Psychol Med 28:551558.

  • Tononi G and Cirelli C (2006). Sleep function and synaptic homeostasis. Sleep Med Rev 10:4962. doi:10.1016/j.smrv.2005.05.002

  • Tucker MA and Fishbein W (2008). Enhancement of declarative memory performance following a daytime nap is contingent on strength of initial task acquisition. Sleep 31:197203.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Tucker MA and Fishbein W (2009). The impact of sleep duration and subject intelligence on declarative and motor memory performance: how much is enough? J Sleep Res 18:304312. doi:10.1111/j.1365-2869.2009.00740.x

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Wechsler D (1987). Wechsler memory scale-revised. San Antonio, TX: The Psychological Corporation.

Electronic Supplementary Material (ESM)

Electronic Supplementary Material (ESM) associated with this article can be found at the website www.akademiai.com/doi/suppl/10.1556/2053.1.2016.001

ESM1 Two study designs

Table S1. Two study designs

Figure S1. Resistance to interference and delayed recall in relation to individual baseline memory performance

ESM2 Videos from 1st International Conference on Sleep Spindling, Budapest, Hungary (May 12–14, 2016)

Individual baseline memory performance and its significance for sleep-dependent memory consolidation – Primary Insomnia; https://youtu.be/-XqE1xSwSss

Individual baseline memory performance and its significance for sleep-dependent memory consolidation – Learning & retrieval; https://youtu.be/9h9AqqtgUYw

Individual baseline memory performance and its significance for sleep-dependent memory consolidation – Interference; https://youtu.be/aIlaJFErLFM

If the inline PDF is not rendering correctly, you can download the PDF file here.

Supplementary Materials

Monthly Content Usage

Abstract Views Full Text Views PDF Downloads
Jun 2020 0 9 2
Jul 2020 0 19 3
Aug 2020 0 6 7
Sep 2020 0 17 3
Oct 2020 0 14 2
Nov 2020 0 19 5
Dec 2020 0 0 0