7. Sleep

Question 1

Jenkins & Dallenbach (1924)

Examined the blip in Ebbinghaus data using more participants

• Between evening and morning (separated by sleep) there seemed to be less forgetting

Answer 1

(learning in the evening vs learning in the morning)

• Found much less forgetting following sleep

!Active role of sleep or lack of interference???

…They argued lack of interference

Question 2

What is polysomnography?

Answer 2

• EEG
  +
• EMG (electromyography - muscle tension, often on chin)
  +
• EOG (electrooculography, eye movements)

Question 3

Characteristics of Sleep:

Stage 1 (shallow sleep)

Answer 3

EEG:

• absence of alpha activity
• vertex sharp waves

EMG:
- relatively low amplitude

EOG:
- Slow eye movement

Question 4

Characteristics of Sleep:

Stage 2

Answer 4

EEG:
- Sleep spindles (oscillating at 12-15Hz)

• K-complexes (high voltage, sharp wave)

EOG:
- no eye movements

Question 5

Characteristics of Sleep:

Stages 3 & 4 (Slow wave sleep - SWS)

Answer 5

EEG:
- slow Delta activity (<=2Hz) with high-voltage (>=75uV)

EMG:
- low tonic activities

EOG:
- no eye movement

“Dead to the world” - very hard to be woken up from SWS

Question 6

Characteristics of Sleep:

REM sleep (rapid eye movement)

Answer 6

EEG:
- mixed frequency (rel low voltage)

EMG:
- ironically suppressed (sleep paralysis)

EOG:
- contains rapid eye movements

Question 7

Plihal & Born (1997):

Partial sleep deprivation

• suggestive evidence for the importance of SWS and REM

Answer 7

Took advantage of half of sleep being SWS and half being REM

Looked at effects on performance in:

1. Standard procedural tasks (mirror tracing):

• People who get REM (late night sleep) do much better in mirror tracing
• 2. Declarative memory tasks (paired-associate learning):
• People who get SWS (or early night sleep) do much better with declarative memory

Question 8

Marr (1970)

System consolidation theory (computational model)

Answer 8

• Hippocampus stores day’s events
• subsequent transfer of info to the neocortex
• transfer based on hippocampal replay

Question 9

Changes in brain representations of memories of pictures (Evidence for system consolidation theory)

Takashima et al (2006)

Answer 9

Longitudinal (3 month) study

Day 1:
1. Study 320 pictures (remote)

2. Rest/nap
3. Study 80 pictures (recent)
4. Recognition test in scanner:
   - 80 remote pics
   - 80 recent pics
   - 80 new (not seen before) pics

Results suggest that:

!SWS is important for declarative memory

!Performance on REMOTE items correlated with hippocampus activity (mediating acquisition of new memories)

!Then these memories taken over by activity in vmPFC

Question 10

ACTIVE Systems Consolidation theory

Born et al (2006, 2010, 2012)

• Modification of standard model (fleshing out sleep’s role)

Answer 10

1. Hippocampus is associated with initial learning of a new declarative memory
2. Sleep is then the medium for hippocampal replay over many nights of sleep
3. Then the consolidated memory becomes ingrained in the cortex without the participation of the hippocampus

Question 11

Systems consolidation of memory:
Rasch & Born (2013)

(Staresina et al, 2015; electrode arrays implanted in the brain)

Role of:

• Slow-wave oscillations
• Spindles
• Hippocampal ripples

Answer 11

Slow-wave oscillations:

• Provide timing signal for other brain components to do their job at the right time (exchange of info between hip and cortex)

Spindles:
- Thalamo-cortical

• Are aligned with S-W oscillations

Hippocampal ripples:
- When spindle wave dips, the ripples are active (synched activity between the two) = Hippocampus sending information to the neocortex?

Question 12

Targeted memory reactivation

Rudoy et al (2009)

If hippocampal replay of memories occurs in sleep, then possibly what you can do is enhance/model this replay synthetically by stimulating the brain

(using sound)

Answer 12

1. Learning objects and their locations (cat, kettle….; declarative memory task)
   - as you see the objects, their characteristic sounds are played along with them (cat + meow)
2. Nap (25 min)
   - played sounds for half of the objects during SWS

At test after the nap:

!!!!Significantly less forgetting for items that were played during sleep

• sound MAY trigger hippocampal replay

Question 13

Marshall et al (2006)

Stimulating the sleeping brain (transcranial DC stimulation)

Answer 13

If you can make the slow oscillations of SWS stronger with DC stimulation that would result in better memory retention for declarative memories

Question 14

Ngo et al (2013)

Enhancing slow waves using sound stimuli (sleep)

Answer 14

Presenting sounds during the up-states (peaks) of slow wave

!!!When stimulated with sounds, the oscillation does not have greater peaks but carries on oscillating for a little longer (~3s)

!!!These longer oscillations led to better retention of information boosted by auditory stimuli in sleep

!!!Also, more SWS = better retention

Question 15

Hippocampal reactivation in animals through electrodes implated into animal brains

(O’Keefe & Nadel, 1978)

place cell activation in a maze

Answer 15

Place cells fire when rats are in a certain location

• likely to be a form of cognitive mapping

Question 16

Hippocampal reactivation in animals through electrodes implated into animal brains

Wilson & McNaughton (1994)

• Slow-wave sleep and reactivation maze navigation

Answer 16

Showed that place cells that fire in sequence when exploring a maze show similar firing sequences in subsequent SWS

(Suggests replay)

! These are also found in nonhuman primates and songbirds

Question 17

Euston et al (2007)

Reactivation in sleep:
- multiple cell recording in mPFC and hippocampus

Answer 17

• found sequences of neurons firing in the mPFC and the hippocampus during maze exploring

!!!in replay during sleep, these sequences fire at a much faster rate

Question 18

Beyond classical sleep consolidation

Walker (2009); Stickgold and Walker (2010)

Sleep is for more than just strengthening/stabilising memory

Answer 18

Unitization:
- Representations initially shared between the hippocampus and the cortex

• then the representation becomes unitized in the cortex as its own thing

Assimilation:

• Changes in the new cortical representation or pre-existing cortical representations to integrate the new representation with other ones

Abstraction:
- Being able to generalize the information like a cortical schema

• You rarely encounter exactly the same circumstances more than once, so this enables a flexibility in applying the established memory

Question 19

Beyond classical sleep consolidation

Kuriyama et al (2004)

Evidence for unitization of representations into the cortex

Answer 19

Learning finger-tapping sequences of 9 elements (e.g. 2, 5, 3, 1, 4, 2…)

They had some evidence that during initial learning we develop “chunks” to break down the sequence

The following sleep, the sequence was unitized (instead of known in chunks, it was known as a whole)

Question 20

Beyond classical sleep consolidation

Tamminen et al (2010)

Evidence for assimilation of representations in the cortex

Answer 20

Lexical assimilation (learning new words) is facilitated by sleep

• learning something new may influence the lexical neighbours of that word

e. g. the “Brexit effect”
- at some point in your life you encountered the word “Brexit” for the first time in your life

• Initially “Brexit” may not be linked with other words (e.g. breakfast~bread brexit)

- In the final state, it may be linked with words that you knew before (e.g. breakfast~bread~brexit)
##################

In the study they taught a new word “Cathedric”

• Similar word such as: Cathedral

Following sleep, the new word started competing with existing words

• This onset of competition was correlated with the amount of sleep spindles present during sleep

(More spindles = bigger changes how new information influenced old information)

• people getting word Brexit and breakfast mixed up after hearing it so much*

Question 21

Beyond classical sleep consolidation

Payne et al (2009)

Evidence for abstraction of representations in the cortex

Answer 21

Hippocampal transfer may lead to reorganisation and restructuring of new knowledge

Used Deese-Roediger-McDermott (DRM) paradigm:
- “bed, rest, awake, tired, dream, wake, snooze, blanket, doze, slumber, snore, nap, peace, yawn, drowsy”

She looked into to what extent people recalled the critical word (Sleep) after period or wake or sleep.

!!!!Following sleep, people were more likely to recall the critical word (Sleep) after sleep compared to wake