Transcript Mobile phone/head interactions

Mobile phone-Human in-vivo studies
Alan W Preece
Medical Physics Research Centre
UNIVERSITY of BRISTOL
Existing Effects
• Pacemakers - (disturbances to the pacing mode
in range 0.1 -2m - but not when implanted)
• ITU - safety critical equipment - MDA guidelines
• Neuro-endocrine - effects on animals, but not humans
•
•
•
•
•
•
VEP - no effect
Subjective effects - Hocking and Hansson-Mild
DNA fragmentation - Lai
Long term potentiation - Tattersall
Blood pressure changes - Braune
Brain permeability - Salford
Selection of key animal studies
• Seaman and Lebowitz ‘89 - acoustic effects of low power
915MHz pulses
• Chou etal ‘85 - no acoustic effect of low power 2.45GHz
pulses
• Thuroczy ‘94 - effects on the EEG and brain blood flow
• Lai and Singh ‘96 - neurotransmitters and DNA
• Salford ‘94 - blood-brain barrier decrease
• Repacholi ‘97 - increased lymphoma incidence
(All below 10W/kg SAR)
Selection of key human studies
• Borbeley et al ‘99 - Effect of GSM on sleep - increase
EEG activity - deeper sleep
• Freude et al ‘98 - Effect of GSM on VEP during visual
tasks (on slow wave activity) - left ear
• Urban et al ‘98 - No effect of GSM on VEP
• Preece et al ‘99 - No effect on memory - but changes in
response speed
• Mild et al ‘98 - Subjective stress and warmth effects (epi.)
• Hardell ‘99 - No increase in absolute no. of tumours, but
association with side of use (recall bias, handedness) (epi)
WTR output - (Summary of letter to Chairman of AT&T
corporation from G Carlo - head of WTR)
• Rate of death from brain Ca. higher in mobile than
portable phone users
• Acoustic neuroma 50% higher in phone users of
>6 years use
• Risk of neuro-epithelial tumours 2x higher in
phone users
• Correlation between side of use and incidence of
tumours
• Confirmation of genetic damage
EEG power density (field % of sham)
120
115
110
105
100
95
0
5
10
15
Frequency (Hz)
20
25
EEG power (field%of sham)
120
10.0 - 11.0 Hz
110
100
90
120
13.5 - 14.0 Hz
110
100
90
0
1
2
3
4
5
6
Time after lights off (h)
7
Waking after sleep onset (min)
Sham
Field
20
16
12
8
4
0
objective
subjective
Waking after sleeping onset (min)
25
20
15
10
5
0
First week field
Field
Sham
First week sham
Plane
wave
Near
field
Field penetration in tissue in near field
Neubar et al. 1999
Antenna
Antenna
Effective
penetration at
900MHz for
human and rat
brain
Thermal effects
• Human cells will stand up to 43C for only
a short time--above this cell killing
becomes very rapid. Probably about 41C
is tolerable.
• To reach this (41o) an SAR of about
10Wkg-1 is needed
• Humans start to sweat at 1-2oC above
normal - about 2-4Wkg-1
Advised levels
Whole body:
Workers - 0.4Wkg-1 (above 10MHz)
Public - 0.08Wkg-1
,,
Localized (including head)
10Wkg-1 NRPB (av. over 10g, 10MHz-10GHz)
or 2Wkg-1 ICNIRP (av. over 10g for general
public)
Both averaged over any 6 min. period
Mobile phones - SAR estimates
Max values (referenced to 1Watt)
Dimbylow (‘94) 3 - 4Wkg-1
Excell (‘98) - 4.2 (900MHz) - 8.2Wkg-1(1.8GHz)
Brain values (referenced to 0.25W GSM)
CSELT (‘99) - 0.3Wkg-1 (Computed standard design)
Measured - 0.44Wkg-1 (Helical and quarter wave ant.)
to 0.02Wkg-1 (Flip-out design)
Phone type
Field strength
in brain
Nokia 5110
18.84
Nokia 2110
20.47
Bosch World 718 17.78
Nokia 6110
16.59
Ericsson GA628 15.88
Hagenuk Global 5.12
Motorola V3688 4.58
Motorola Star Tac 70 4.56
SAR (W/kg)
in brain
0.37
0.44
0.33
0.29
0.26
0.03
0.02
0.02
Summary
Highest brain SAR from GSM (0.25W) = 0.44Wkg-1
Range of SAR from analogue = 0.9 - 1.76Wkg-1
Thermal estimates
Worst case for head:
4Wkg-1 for 30mins = 7200J = 1714 calories
that could be 1.7o
Worst case for brain:
0.44Wkg-1 for 30min - 0.20
1.6Wkg-1 for 30min - 0.70
With blood flow, half these values or less are to be expected
(Rui and Foster 1999)
Neubar et al. 1999
Dielectric property of head tissues
Tissue Type
Fat
Skull
Skin
Brain
Eye
Muscle
(Air-way)

5.5
12.5
39.5
56.8
52.3
55.0
(1)

1.8
3.4
14.0
22.0
23.2
22.0
(0)
Human Cognitive Function Tests
Standard tests used to assess effects of drugs e.g. opiates,
psychotropics for medico-legal purposes
1. Immediate Word Recall
2. Picture Presentation
3. Simple Reaction Time
4. Digit Vigilance Task
5. Choice Reaction Time
6. Spatial Working Memory
7. Numeric Working Memory
8. Delayed Word Recall
9. Delayed Word Recognition
10. Delayed Pict. Recognition
Effect at 50Hz - 600T applied to
the head (about 6mA/sq.m)
•
•
•
•
•
•
•
Reaction time - no effect
Choice reaction - no effect
Spatial memory - no effect
Vigilance - no effect
Picture recall - small effect
Number recall - intermediate effect
Word recall - large effect
Simulated 915MHz phone
Exposure
•1Watt CW to antenna, or
•Modulated at 217Hz square wave at 0.125 duty
cycle (mean power 0.125 W)
•Mounted on ear defender with antenna 2-3cm from
head, over left squamous temple bone
Volunteers
Group 1. Hospital staff aged 24 to 60
No control over alcohol, caffeine, sleep, drugs
Group 2. Medical, dental or postgrad. science students aged
18-28
Controlled for alcohol, caffeine, sleep, drugs
All were given two training sessions, and 30mins exposure
each visit
To
Coding
switch
217
Hz
915
MHz
Result 1:
• Group 1 showed no effect on memory
but “Choice reaction time” was decreased i.e. the speed went up!!
– is this a Type 1 error?
– or, did some volunteers have no sleep, too
much alcohol or too much coffee?
Reason for the second, tightly controlled, group
Result 2:
• Second group showed exactly the same
result!
– No effect of alcohol (even after 15 units the
night before)
– Coffee/tea uniformly distributed
– Sleep unrelated to performance
– Drugs (legal or otherwise) not a factor
Variable
(1st four)
P-Value
(3-way
comparison)
0.559
Simple
reaction time
(msec)
Choice
0.003
reaction time
(msec)
Vigilance
0.377
reaction time
(msec)
Vigilance
0.164
accuracy (%)
Mean
Analogue
field
225.9
Mean
Digital
field
228.6
Mean
Cont
conditions
227.3
373.4 384.4 387.9
385.8 382.2 388.1
99.0
%
99.1
%
98.3
%
Results (MANOVA):
p-value for 3
simulations
0.007
p-value for
visit order
NS
NS
0.000
Accuracy on working
and secondary
memory tasks
NS
NS
Accuracy on
attentional tasks
NS
NS
Outcome
variable
Reaction time on
attentional tasks
Speed on working
& secondary
memory
Conclusions:
• There is an effect on the brain of a 915MHz
transmission
– Greater for analogue than GSM type
– mean effect is a 15msec reduction in transit
between visual and speech centre - this has to
be a synaptic effect not transmission one
– Big question is whether this is thermal or nonthermal
Possibilities:
• An effect on the angular gyrus - this is a
processing centre between visual and
speech centres
– Normally on the left
– Is affected by heat or blood flow - (measurable)
– Could be protein effect - (effect will persist)
These are all testable hypotheses
Brain blood flow measurement methods
MRI - difficult without contrast, but Diffusion-perfusion
sequence possible (extra large imaging gradient). Metal
antenna a nuisance
PET - Involves radioactive materials (need about 4mSv) and
specialised centre to carry it out, and expensive
Ultrasound - Direct colour-Doppler studies on the middle
cerebral artery which perfuses the cortex
NIRS - transillumination of the skull and brain
Flow measurement protocol
Dual channel 2Mhz ultrasound colour Doppler with timeaverage-max and resistance-index measurement
Two dummy phones on right and left side, one activated with
915Mhz analogue to give SAR of 1WKg-1 in brain
Ten volunteers on two visits each, randomly assigned to right
or left activation, contralateral side on second visit
Record 15 min baseline flows on right and left
Record 15 min with one phone activated
Record 15 min recovery
Minutes
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0
TAM and RIx100
Blood flow / 2Watts/kg
70
RI.
60
50
TAM
40
Flow
30
20
Power "on"
10
0
Possible mechanism via HSP
• Brophy (1998) - HSP in intimal epithelium
causes vasodilatation
– Used to precondition vessels before vascular
surgery
• If waves can induce HSP in absence of
heat then this may explain vasodilatation at
very low RF levels - (Chronic effects?)
Blood flow
A change in blood flow in the middle cerebral artery
can be detected at power levels 2 - 3x usual phone
power
Similar experiments with NIRS show changes in
perfusion (but are these SKIN and SKULL changes
only)
Changes in BRAIN perfusion may be more difficult
to pin down - in which case are they even
significant???
Summary and Conclusions
• Some evidence for a direct electro-physiological
effect on brain of pulsed RF (i.e.GSM systems)
• Some evidence for short term physiological effects
(perfusion, temperature)
• Some animal and cellular evidence for heat-shock
(stress) protein at non-thermal levels
• Confusing evidence for genotoxic damage
• No sound evidence as yet for human
carcinogenicity
Needs for future study
• Epidemiology of cognitive changes (i.e. long term
memory effects)
• Epidemiology of neural tumours (already under way
with IARC)
• Correlation of SAR’s from phones and “normal” limits
for physiological changes in the brain
• Study of other critical organ exposure (salivary,
thyroid, eye)
• Resolution of the genotoxicity (Lai and Singh) issue and
confirmation of stress protein production
• New technology - Tetra, satellite, environment
Science and Technology Select Committee
Report - September 1999
•
•
•
•
•
•
•
•
•
Change limits to ICNIRP (2W/kg instead of 10W/kg)*
Set up Independent Expert Group
Industry should be involved, but in minority
Lay members on study groups
Confirm or deny work on DNA fragmentation*
Further research on subjective and cognitive symptoms
Large increase in publicly-funded research
Academically led research programme
Publish SAR’s and give consumers choice*