Está en la página 1de 33

Interphone Studies To Date

An Examination of Poor Study Design Resulting in an UNDER-ESTIMATION of the Risk of Brain Tumors
L. Lloyd Morgan BEMS, San Diego, 12 June 2008

L. Lloyd Morgan [bilovsky@aol.com]

Methodology
What If There Is No Risk of Brain Tumors?
ORs <1.0 would be ~equal ORs>1.0
Think coin tossing
OR=1.0 are excluded

~5% of ORs would be significant


ORs <1.0 would be ~equal ORs>1.0

Calculate ratio: OR<1.0/OR>1.0


13 Interphone brain tumor studies to date
Exclude 2 overlapping studies and recent (Schlehofer) study Analysis restricted to 10 Interphone brain tumor studies

Calculate binomial p-values


L. Lloyd Morgan [bilovsky@aol.com] 2

Methodology
Calculate Ratio by Categories by Studies

How to have statistically independent categories?


Compare between studies, not within studies Categories
Brain Tumors
All Acoustic Neuroma Glioma Meningioma

Years of use (Years) Cumulative hours of use (Hours) Cumulative number of calls (Call #) Regular cellphone use (Regular) Years of ipsilateral cellphone use (Years Ipsi) Years of contralateral cellphone use (Yrs Contra) Minutes of cellphone use per day (Min/Day)
L. Lloyd Morgan [bilovsky@aol.com] 3

Results
Percent Significant Findings By Category
Expectation: ~5%
25%

20%

15%

10%

~ expected
5%
Based on 7 Findings 0%

0% All Years Hours Call # "Regular" Years Ipsi Years Contra

Min per Day

Categories
L. Lloyd Morgan [bilovsky@aol.com] 4

Results
Ratio (OR<1.0/OR>1.0): Exposures: >10 Year and <10 Year
Ratio

5.0
p=3.6 x 10-22

4.0

3.0

2.0
p=0.16
~ Expected ratio if no risk

1.0

0.0

>10 year
Longest Exposure Lowest Ratio
L. Lloyd Morgan [bilovsky@aol.com]

<10 year

Results
Ratio by Category
Ratio

8 7 6 5 4 3 2 1 0
All Call # Hours Years "Regular" Years Contra Min per Day Years Ipsi

p=6.0x10-07

Non-significant Near-significant

p=6.1x10-06 p=1.2x10-20 p=1.1x10-07 p=0.0011 p=0.0097 p=0.16 p=0.097

Categories

Highest Exposure Lowest Ratio

L. Lloyd Morgan [bilovsky@aol.com]

Results
Ratio by Brain Tumor Type
Ratio

6 5
p=1.2x10-20 p=6.0x10-10

p=8.2x10-9

4 3 2 1 0 All Acoustic Neuroma Glioma Meningioma


p=2.9x10-5

L. Lloyd Morgan [bilovsky@aol.com]

Interphone Protocol Design Flaws


Flaw 1: Selection Bias
Participating controls use cellphones more than nonparticipating controls
Weighted average control participation rate: 59% Lon 2004: 20% control refused; 34% used, 59% did not use

Underestimates risk

Flaw 2: Tumors outside the radiation plume are unexposed


Unexposed tumors treated as exposed Plume volume small relative to brain volume
Well know since 1994 (4 previous papers)

Underestimates risk
L. Lloyd Morgan [bilovsky@aol.com] 8

Flaw 2 Tumors Outside Radiation Plume Are Unexposed


Radiation plumes volume is small % of brains volume Ipsilateral: exposed Contralateral: unexposed
Absorbed radiation decreases rapidly with plume penetration depth
Half-way to the brains mid-line, >90% of energy is absorbed

Percentage of absorbed cellphone radiation


Ipsilateral temporal lobe:
~15% of brains volume

50-60% (wt. av.=53%) 12-25% (wt. av.=19%)

Ipsilateral cerebellum:
~5% of brains volume

62-85% of absorbed radiation is in ~20% of the brains volume


Plume decreases rapid with depth (actual exposed brains volume: <20%, perhaps 15%)
L. Lloyd Morgan [bilovsky@aol.com] 9

Flaw 2 Absorbed Radiation Decreases Rapidly w Depth


Relative Absorbed Radation and Penetration Depth in Temporal Lobe % Absorbed Radiation
Relative to Max Absorbed Radiation

60%
900 MHz European Phones (worst case)

50% 40% 30% 20% 10% 0%


15-24
Source: Cardis et al 2008

800-900 MHz Japanese Phones

25-34

35-44

45-54

55-64

65-74

75-84

Depth (mm)

L. Lloyd Morgan [bilovsky@aol.com]

10

Interphone Protocol Design Flaws


Flaw 3: Latency time
Known latency times
Ionizing radiation & brain tumor: 20-40 years Smoking & lung cancer: ~30 years Asbestos & mesothelioma: 20-40+ years

Short latency times underestimates risk

Flaw 4: Definition of regular user


Regular user: At least once a week for 6 months or more
If definition of regular smoker were used, would a risk of lung cancer be found?

Definition of regular user underestimates risk


L. Lloyd Morgan [bilovsky@aol.com] 11

Flaws 3 and 4 Latency Time and the Definition of Regular Users


UK Subscribers by Year
Millions

60 50 40 30
>5 year latency 15% User-years

20
>10 year latency 2% User-years

<5 year latency 85% User-Year

10 0

1 2 3 4 5 6 7 15 14 13 12 11 10 9

8 8

9 7

10

11

12

13

14

Years from Eligibility Date (Latency Time)


L. Lloyd Morgan [bilovsky@aol.com]

1 16 0 Wt. Ave. Eligibility Date 2002.5


15

12

Interphone Protocol Design Flaws


Flaw 5: Young adults and children excluded
Children and young adults at greater risk than adults
Interphone Protocol: 30-59 years
Some studies reduce minimum age to 20 years

Underestimates risk

L. Lloyd Morgan [bilovsky@aol.com]

13

Flaw 5 Young Adults and Children Excluded


Sweden: Cellphone Rad.
OR

Korea: Cellphone Rad.


OR

Increased Risk of Brain Tumor

4
Source: J.W. Choi el al.Case-control Studies on Human Effects of Wireless Phone RF in Korea, BEMS 2006

8 7 6 5 4 3 2 1 0
20-80 years Analog cellphone

P<0.01
Source: Hardell et al. Arch Environ Health. 2004 Mar;59(3):132-7.

20-29 years

20-80 years

20-29 years
0
20-29 years 30-39 years
Age Range

Cordless phone

40-49 years

50-59 years

Israel: Ionizing Radiation


ERR/GY

Excess Relative Risk (ERR) per Gray (Gy) Malignant Brain Tumors by Age from Ionizing Radiation Exposure

400% 356% 300%


Mean estimated dose: 1.5 Gy (range 1.0 to 6.0 Gy)

200%

224%

100%

47% 0% <5 5-9 10+

Age at Exposure
Source: Sadetzki et al., RADIATION RESEARCH 163, 424432 (2005)

L. Lloyd Morgan [bilovsky@aol.com]

14

Interphone Protocol Design Flaws


Flaw 6: Comparison cellphone radiated power: higher vs lower
Analog Vs Digital phones
No longer possible

Rural Vs Urban users Underestimates risk

L. Lloyd Morgan [bilovsky@aol.com]

15

Interphone Protocol Design Flaws


Flaw 7: Cordless phone, walkie-talkie, Ham, and proximity to TV & radio transmitters
Treated as unexposed Underestimation of risk

Flaw 8: Exclusion of brain tumor types


Includes only acoustic neuroma, glioma & meningioma Other brain tumor types are excluded
For example lymphoma and neuroepithelial brain tumors

Underestimates risk

Flaw 9: Exclusion of brain tumor cases because of death


Underestimates risk of most deadly brain tumors
L. Lloyd Morgan [bilovsky@aol.com] 16

Interphone Protocol Design Flaws


Flaw 10: Recall bias
Light users underestimate use Heavy users overestimate use Result: Large underestimation of risk

L. Lloyd Morgan [bilovsky@aol.com]

17

How to Resolve Flaws


Increase diagnosis eligibility time
Nine Interphone studies: weighted-average 2.6 years
Hardell et al. eligibility time: 6 years

Lower age range to <15 years Pay controls (and cases?) for participation in study
Do not tell controls what is the purpose of the study

Interview proxies in case of death


Separately report both case and proxy interview results

Treat unexposed tumors as unexposed Etc., Etc., Etc.,


L. Lloyd Morgan [bilovsky@aol.com] 18

Conflict-of-Interest
Cellphone Industry
Interphone funding is inadequate to resolve flaws
More funding, greater potential of substantial revenue loss

Researchers conflict-of-interest (unconscious?)


Source of funds is known in spite of Firewall Honest, but Dont bite the hand that feeds you
90 significant protective results
Ignored by authors (no commentary in the text)
L. Lloyd Morgan [bilovsky@aol.com] 19

Potential Brain Tumor Risk 30-year Latency


Poisson Distribution Calculation
Risk

100%
1985 1st Use 1995 10 Yrs 2005 20 Yrs 2015 30 Yrs

97%

80%

60%
55%

40%

20%
4%

0%
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40

0.002%

Years Since First Exposure

L. Lloyd Morgan [bilovsky@aol.com]

20

Potential Public Health Risk


Potential Brain Tumor Cases From Use of a Cellphone Potential Cases Assuming a 30-Year Latency Time and 10% of Users1 of Brain Tumors per Year Diagnosed with a Brain Tumor
1,800,000 1,600,000 1,400,000 1,200,000 1,000,000 800,000 600,000 400,000 200,000 0
1

Cellphone Subscribers millions 450 400 350

Year 2019 1,590,513

Based on 10% of long-term smokers are diagnosed with lung cancer


Source Cellphone Subscribers: CTIA Source brain tumor diagnosed in 2004: CBTRUS

300 250 200 150

Year 2004 44,447 Dx ~3,600 from cellphone use

100 50 0

19 8 19 5 8 19 6 8 19 7 8 19 8 8 19 9 9 19 0 9 19 1 9 19 2 9 19 3 9 19 4 9 19 5 9 19 6 9 19 7 9 19 8 9 20 9 0 20 0 0 20 1 0 20 2 0 20 3 0 20 4 0 20 5 0 20 6 0 20 7 0 20 8 0 20 9 1 20 0 1 20 1 1 20 2 1 20 3 1 20 4 1 20 5 1 20 6 1 20 7 1 20 8 19

L. Lloyd Morgan [bilovsky@aol.com]

21

Conclusions
Interphone results substantially underestimate the risk of brain tumors
Great majority of results have OR<1.0
Either cellphone use is protective, or the study has major flaws

Ratio is lowest for highest exposures: ipsilateral use or >10 years of use Significant risk found for >10 years and ipsilateral use

Without design flaws Odds Ratios would increase substantially Cellphone industrys conflict-of-interest is obvious Government: ignores potential epidemic (see no evil)
Public health impact is enormous

Industry independent studies are required


L. Lloyd Morgan [bilovsky@aol.com] 22

I Pray Im Wrong!

L. Lloyd Morgan [bilovsky@aol.com]

23

Now What?
Based on CBTRUS Incidence Data

Window closed for case-control studies


No unexposed cases remain

Cohort studies
Unable to know users of company owed cellphones Unable to interview cellphone users Requires enormous numbers
1,000,000 user-years will find (assuming cellphones do not increase risk)
~6 acoustic neuromas ~54 gliomas ~45 meningiomas

Requires ~1 billion user-years to analyze by


Gender, SES, Years of use ,Exposed tumors only

Requires 30 year cohort study


L. Lloyd Morgan [bilovsky@aol.com] 24

Interphone Protocol Design Flaws


Flaw 11: Recall bias
Interview cases immediately after diagnosis and 6 months after surgery
Improved memory and cognition 6 months after surgery

Flaw 12: Observational bias


Interviewer not blinded with face-to-face interviews
Mailed questionnaires provide blindness
Supplement by phone as necessary

Flaw 13: Too few cases for statistical power


Nine Interphone Brain Tumor Studies: Use for >10 years
Average 18 cases per study At minimum requires 2-fold more cases and controls for sufficient statistical power
L. Lloyd Morgan [bilovsky@aol.com] 25

Design Changes to Resolve Flaws


Treat unexposed tumors as unexposed
Tumors outside radiation plume
Data was available, but to date not used, or even discussed
Too few cases?

Treat RF/MW exposures and exposed


Cordless phone, walkie-talkie radios, Ham transmitters

Overweight rural users or increase eligibility time


Compare risk of brain tumor with rural and urban users
Requires sufficient number of cases and controls

Use questionnaires not face-to-face interviews


L. Lloyd Morgan [bilovsky@aol.com] 26

Design Changes to Resolve Flaws


Reporting regular use Do not publish regular use data At minimum report regular use for >5 years, or >10 years
Assumes >3-fold increase in case eligibility range

Latency time: initiation or promotion? Some researchers assume cellphone can only be promoters
What is evidence for initiation vs promotion?

Follow cases & controls for a longer period


L. Lloyd Morgan [bilovsky@aol.com] 27

Design Changes to Resolve Flaws


Increase eligibility time to 9 years (for sufficient statistical power)
>3-fold increase in cases and controls Publish results every 3 years Provides longer latency time Resolves whether cellphones use initiates or promotes tumors

L. Lloyd Morgan [bilovsky@aol.com]

28

Flaw 2 Tumors Outside Radiation Plume Are Unexposed


900 MHz European Phone: Depth by Structure
Relative SAR

60% 50% 40% 30% 20% 10% 0% 15-24 25-34 35-44 45-54 Depth (mm) 55-64 65-74 75-84

50%

Total Raditaion Absobed by Structure


19% 9% 5% Frontal Parietal Occipital

12%

Temporal

Cerebellum

L. Lloyd Morgan [bilovsky@aol.com]

29

Flaw 5 Children Excluded Ionizing Radiation Example


ERR/GY

Excess Relative Risk (ERR) per Gray (Gy) Malignant Brain Tumors by Age from Ionizing Radiation Exposure

400% 356% 300%


Mean estimated dose: 1.5 Gy (range 1.0 to 6.0 Gy)

200%

224%

100% 47% 0% <5 5-9 Age at Exposure 10+

Source: Sadetzki et al., RADIATION RESEARCH 163, 424432 (2005)

L. Lloyd Morgan [bilovsky@aol.com]

30

Flaw 5
OR

Young Adults Excluded Korean Cellphone Study


P<0.01
Source: J.W. Choi el al.Case-control Studies on Human Effects of Wireless Phone RF in Korea, BEMS 2006

0
20-29 years 30-39 years
Age Range

40-49 years

50-59 years

L. Lloyd Morgan [bilovsky@aol.com]

31

Flaw 5
Young Adults Excluded Swedish Cellphone Study
OR

Increased Risk of Brain Tumor

8 7 6 5 4 3 2 1 0
20-80 years Analog cellphone

Source: Hardell et al. Arch Environ Health. 2004 Mar;59(3):132-7.

20-29 years

20-80 years

20-29 years

Cordless phone
32

L. Lloyd Morgan [bilovsky@aol.com]

Flaw 2 Tumors Outside Radiation Plume Are Unexposed

~5.6 cm

Source: http://serendip.brynmawr.edu/bb/kinser/Glossary.html

~10.4 cm Surface area ~162 cm2

Ipsilateral temporal lobes volume to total brains volume ~15% Ipsilateral cerebellums volume to total volume ~5% Worst case: ~62% of ipsilateral radiation is absorbed in ~20% of brain
L. Lloyd Morgan [bilovsky@aol.com] 33