Air Force Health Study Update 10 June 2005

1
Air Force Health Study Update10 June 2005

• Presenter
• Joel Michalek, PhD
• Air Force Research Laboratory
• Brooks City-Base, Texas

2
Contents

• Overview of the 2002 report
• Research
• Diabetes
• Cancer
• Reported sleep abnormalities
• Interactions with the IOM

3
Overview of the 2002 Report
4
Methods

• Summarized health conditions observed at the 2002
  AFHS physical examination (N1951, Ranch Hand
  777, Comparison 1174)
• Main effects linear, logistic, and proportional
  hazards models with covariate adjustment
• Data sources
• Laboratory
• Questionnaire
• Physical examination
• Medical records

5
Results

• General health
• Body mass index increased with 1987 dioxin
• Neoplasia
• A mixed pattern of associations with no
  suggestion of internal or external consistency
• Neurology
• The risks of abnormal pin prick (not bilateral)
  and absent patellar reflex were increased in the
  High dioxin category

6
Summary

• Psychology
• No outcome measures were associated with
  herbicide or dioxin exposure
• Gastrointestinal
• The risk of abnormally high triglycerides was
  increased in Ranch Hand enlisted ground crew and
  the High dioxin category

7
Summary

• Dermatology
• The frequency and occurrence of reported acne
  after service in SEA was increased in Ranch Hand
  enlisted ground crew, and in the Background, Low,
  and High dioxin categories
• The duration of reported acne after service in
  SEA increased with 1987 dioxin
• The frequency of acneiform lesions was increased
  only in the Background dioxin category
• Secondary skin lesions were not associated with
  herbicide or dioxin exposure

8
Summary

• Cardiovascular
• Associations observed were not consistent or
  clinically interpretable as adverse
• Hematology
• Associations observed did not suggest an adverse
  relation between herbicide or dioxin exposure and
  any hematological diagnosis
• Renal
• The few associations observed did not indicate an
  adverse relation between renal function and
  herbicide or dioxin exposure

9

• Endocrine
• The risk of diabetes requiring insulin control
  was increased in the High dioxin category
• Fasting insulin and the risk of diabetes
  requiring insulin control increased and time to
  diabetes onset decreased with initial dioxin
• The risk of diabetes requiring oral hypoglycemic
  or insulin control and the risk of diabetes
  requiring insulin increased with 1987 dioxin
• The risk of abnormally high hemoglobin A1c
  increased with dioxin

10

• Endocrine
• Associations between herbicide or dioxin exposure
  and measures of thyroid function or gonadotropins
  did not appear consistent or clinically important
• The adverse associations between type 2 diabetes
  and dioxin exposure were consistent with findings
  at previous physical examinations

11

• Immunology
• There was no consistent or interpretable
  association between any measure of immune
  function and herbicide or dioxin exposure
• Pulmonary
• The associations observed did not suggest an
  adverse relation between respiratory health and
  herbicide or dioxin exposure

12
Diabetes, Dioxin, Calendar Period of Service,
and Days of Spraying

• Another Look at the Check Mark Pattern

13
Diabetes

• Background
• Diabetes risk increased, time to diabetes onset
  decreased, and diabetes severity increased with
  dioxin (Henriksen et al, Epidemiology
  19978(3)252-258)
• Dioxin category analysis produced a check mark
  pattern
• Statistical modeling of the check mark pattern
  was attempted but not published

14
Diabetes

• New analyses of dioxin and diabetes were
  conducted with adjustment for
• Days of spraying
• Calendar period of spraying
• Two rounds of analyses were conducted
• Cycle 5
• Cycle 6
• Time to diabetes onset measured from end of
  qualifying tour

15
Cumulative Days of Spraying by Last Ranch Hand
Year
16
Sample Reduction
Up to and including cycle 5
17
Demographics by Calendar Period1
1. Up to and including cycle 5. 2. During or
prior to 1968
18
Demographics by Spraying Category
19
Ranch Hand Dioxin by Calendar Period of Service

• 1. During or prior to 1969, 2. Log units, 3.
  Yes-No, 4. Adjusted for age, BMI at tour, last
  year in Vietnam.

20
Ranch Hand Dioxin by Spray Category

• 1. High90 d, 2. Log units, 3. Yes-No, 4.
  Adjusted for age, BMI at tour, last year in
  Vietnam.

21
Diabetes by Groupwithout Stratifying

• Adjusted for BMI at blood draw, pk-yr82, fh diab,
  BMI tour, yob, last year, VNdays/SEAdays,
  occupation

22
Diabetes by Dioxin Categorywithout Stratifying

• Adjusted for BMI at blood draw, pk-yr82, fh diab,
  BMI tour, yob, last year, VNdays/SEAdays,
  occupation

23
Diabetes by Groupand Calendar Period of Service

• Adjusted for BMI at blood draw, pk-yr82, fh diab,
  BMI tour, yob, last year, VNdays/SEAdays,
  occupation

24
Diabetes by Groupand Spray Category

• Adjusted for BMI at blood draw, pk-yr82, fh diab,
  BMI tour, yob, last year, VNdays/SEAdays,
  occupation

25
Diabetes by Dioxin Categoryand Calendar Period
of Service

• Adjusted for BMI at blood draw, pk-yr82, fh diab,
  BMI tour, yob, last year, VNdays/SEAdays,
  occupation

26
Diabetes by Dioxin Categoryand Spray Category

• Adjusted for BMI at blood draw, pk-yr82, fh diab,
  BMI tour, yob, last year, VNdays/SEAdays,
  occupation

27
Diabetes by Dioxin Categoryand Calendar and
Spray Category

• Adjusted for BMI at blood draw, pk-yr82, fh diab,
  BMI tour, yob, last year, VNdays/SEAdays,
  occupation

28
Conclusions

• These data support the hypothesis that herbicides
  were more heavily contaminated with dioxin early
  in the war
• Days of spraying and calendar year of service are
  effect modifiers in the association between
  dioxin and diabetes
• The check mark pattern is an artifact of lack of
  adjustment
• Other unknown factors may be important

29
Cancer, Dioxin, Calendar Year of Service, and
Days of Spraying

• Adjusting for Additional
• Operational Factors

30
Background

• Associations between cancer and dioxin are not
  evident in simple main effects models
• Years of service in Southeast Asia is a risk
  factor for cancer in Comparisons
• Cancer vs dioxin associations were revealed in
  Ranch Hands after stratification by
• Years of service in SEA
• The proportion of time spent in Vietnam
• Calendar period of service and days are spraying
  are considered here

31
Methods

• Cancer determined through 30 September 2004
• Onset to the first diagnosis of SEER cancer
  measured from 1 January 1982
• Time to onset for veterans without cancer was the
  date of last physical plus 2 years or date of
  death
• Cancer was determined by review of medical
  records and death certificates
• Proportional hazards models for time to cancer
  onset adjusted for year of birth, military
  occupation, skin reaction to sunlight, eye color,
  smoking history from baseline (pack-years), years
  served in Southeast Asia

32
Methods

• Veterans compliant to at least one physical
  examination were included

33
Sample Reduction

• Fully or partially compliant to any physical
  examination
• Not fully compliant to at least one physical
  examination

34
Cumulative Days of Spraying by Last Ranch Hand
Year
35
Strata

• Early service and increased spraying
• Ranch Hand Service in the Ranch Hand Operation
  during or prior to 1968 and at least 30 days of
  spraying during or prior to 1967
• Comparison Service in Southeast Asia during or
  prior to 1968
• Served at most 2 years in Southeast Asia

36
Analysis Sample Sizes1

• White. Pre-baseline cancer, Ranch Hands with
  missing dioxin
• Excluded (N2262).

37
DemographicsEarly Service and Increased Spraying
38
DemographicsComplement Stratum
39
DemographicsEarly Service and Increased Spraying
40
DemographicsComplement Stratum
41
Ranch Hand Dioxin by Calendar Period of Service

• 1. During or prior to 1968 and 30 pre 67 spray
  days , 2. Log units, 3. Yes-No, 4. Adjusted
  for age, BMI at tour, last year in Vietnam.

42
Ranch Hand Dioxin by Spray Category
1. Adjusted for age at tour, BMI at tour, last
year of Rand Hand service
43
SEER Cancer Unrestricted, White

• Adjusted for year of birth, military occupation,
  smoking history (pack-years),
• skin reaction to sun, eye color, years in SEA

44
SEER Cancer 30 pre-1967 spray days,
Service1968, White
1. Adjusted for year of birth, military
occupation, smoking history (pack-years), skin
reaction to sun (burns painfully or freckles with
no tan, burns or tans mildly, tans deep brown) ,
eye color (brown, hazel/green and gray/blue),
years in SEA
45
SEER Cancer Complement, White
1. Adjusted for year of birth, military
occupation, smoking history (pack-years), skin
reaction to sun, eye color, years in SEA
46
SEER Cancer 2 Years in SEA, White
1. Adjusted for year of birth, military
occupation, smoking history (pack-years), skin
reaction to sun, eye color, years in SEA
47
SEER Cancer 30 pre-1967 SprayDays,
Service1968, 2 Years in SEA, White
1. Adjusted for year of birth, military
occupation, smoking history (pack-years), skin
reaction to sun, eye color, years in SEA
48
SEER Cancer by Spray Category, White
1. Adjusted for year of birth, military
occupation, smoking history (pack-years), skin
reaction to sun, eye color, years in SEA
49
SEER Cancer vs Spray 30 pre-1967 SprayDays,
Service1968, 2 Years in SEA, White
1. Adjusted for year of birth, military
occupation, smoking history (pack-years), skin
reaction to sun, eye color, years in SEA
50
Conclusions

• These data are consistent with the hypothesis
  that herbicides were more heavily contaminated
  with dioxin in the early years of the war
• The number of spray days correlates with dioxin
  in Ranch Hand veterans
• Calendar period of service correlates with dioxin
  in Ranch Hand veterans
• Calendar period of service, spray days, and years
  served in SEA are effect modifiers of the
  relation between dioxin and SEER cancer

51
Reported Sleep Disorders and Dioxin in Veterans
of Operation Ranch Hand

• Liu1 Y, Michalek2 JE, Lian1 M,
• Dickerson1 RL, Frame1 LT
• Texas Tech University, Lubbock, Texas
• Air Force Research Laboratory, San Antonio, Texas

52
Background

• Sleep disruption is one of the most common
  complaints cited by chemical plant workers with
  occupational exposure to dioxin
• Sleep deprivation is associated with glucose
  intolerance, type 2 diabetes, cardiovascular
  disease, and memory deficits
• This is the first epidemiological study of
  reported sleep quality and serum dioxin
  measurements

53
Methods

• Reported sleep problems were obtained during
  interviews in 1987 and 1992
• Each participant was asked 11 questions to assess
  sleep quality, based on a study of the prevalence
  of sleep disorders conducted in Los Angeles in
  1979
• Veterans compliant to the 1987 or 1992 AFHS
  physical examinations were included
• Statistical analysis was based on contrasts of
  Ranch Hands in the Background, Low, and High
  dioxin categories with Comparisons

54
Methods

• Insomnia was defined by the occurrence of at
  least one of
• Having trouble falling asleep
• Waking up during the night
• Waking up too early and cannot go back to sleep
• Waking up unrefreshed

55
Methods

• Parasomnia was defined by the occurrence of at
  least one of
• Frightening dreams
• Sleep talking
• Sleep walking
• Abnormal movement or activity during the night

56
Methods

• Main effects logistic regression models were
  used, adjusted
• Age
• Race (black, non-black)
• Education (high school, college)
• Total household income (0-55k, 55k)
• Marital status (yes, no)
• Alcohol consumption (drink-years) (0,0-40,40)
• Cigarette smoking (pack-years) (0,0-10,10)
• Body mass index at the time of service in
  Southeast Asia

57
Sample Reduction
58
Dioxin Categories
59
1987 Demographics
60
Sleep Complaints in 1987 and 1992
1. Significantly increased relative to
Comparisons (p 61
1987 Sleep Disturbances ()
1. OR3.1, 95 CI 1.6-6.0
62
1992 Sleep Disturbances ()
1. OR1.4, 95 CI 1.1-1.8
63
1987 Sleep Disturbancesand Dioxin Category ()

• OR3.5, 95 CI 1.6-7.8
• OR2.9, 95 CI 1.3-6.3

64
1992 Sleep Disturbancesand Dioxin Category ()
1. OR1.4, 95 CI 1.01-1.90
65
Insomnia Trajectories
1. Significantly different from Comparisons
(p 66
Other findings

• The odds of any sleep problem in 1987 were
  increased in Low and High Ranch Hands relative to
  Background Ranch Hands
• The odds of any sleep complaint in 1992 were
  increased in Ranch Hand Low and High category
  relative to Comparisons
• Odds of snoring and disabling daytime fatigue
  increased with dioxin in Ranch Hands in 1987 but
  not 1992

67
Conclusions

• The increased odds of snoring and disabling
  fatigue and the persistence of insomnia in Ranch
  Hands with elevated dioxin suggest that low-dose
  dioxin exposure targets systems of arousal
  regulation
• The increased risk of snoring and disabling
  fatigue may account for diminished sleep quality
  among Ranch Hands with elevated dioxin

68
Interactions with the IOM

• We briefed the IOM committee in February on
• Study history
• Study design
• Methods of analysis
• Materials
• Ongoing research
• An IOM subcommittee will visit the Air Force
  Health Study facilities on 27 May 2005