Gregory C. Gray, MD, MPH,1 Troy McCarthy,1 Ana W. Capuano, MPS, 1 Charles F. Lynch, MD, PhD,2 Debbie

1
Gregory C. Gray, MD, MPH,1 Troy McCarthy,1 Ana W.
Capuano, MPS, 1 Charles F. Lynch, MD, PhD,2
Debbie A. Wellman,1 Kelly A. Lesher,1 Sharon F.
Setterquist, MT(ASCP),1 Norma J. Miller,1
Patricia A P Gillette,MPH,2 Christopher W. Olsen,
DVM, PhD,3 Alexander I. Klimov, PhD,4 Michael C.
Alavanja, PhD,5 and Jackie M. Katz, PhD6
Background - Swine may play an important role in
cross-species influenza transmission and the
genesis of novel influenza strains. We
prospectively studied swine workers for
serological evidence of zoonotic influenza
infection. Methods - Using data from the
Agricultural Health Study (AHS), we identified,
screened, and enrolled 803 rural Iowans from 29
counties in a 2-year prospective zoonotic
influenza transmission study. Enrollment sera
were studied with hemagglutination inhibition
assays against two swine and one human influenza
viral strains. Demographic and occupational risk
factors were studied for associations with
antibodies against swine H1N1 and H1N2 viruses
using a multivariate proportional odds model at
enrollment (AHS and Non-AHS) and follow-up (AHS
only). Results - Among the 803 subjects, 707
worked in swine production (AHS-E), 80 reported
no previous swine exposure (AHS-NE) and 15 were
excluded because of reported accidental
self-injection with swine vaccine. Data and sera
from 79 non-AHS control (N-AHS) subjects from a
parallel study were used as a reference. (71) of
the 80 AHS-NE live in the same household (mostly
wives) of AHS-E. Multivariate models controlled
for human influenza viruses, flu shot, age,
and/or gender. At enrollment, the AHS-E (10 or
10 years of exposure) and the AHS-NE had
markedly elevated adjusted odds of increased
antibody titers against swine H1N1 (OR, 95CI
55.6, 13.1-236.151.3,11.1-236.5 28.2,
6.1-130.1) and swine H1N2 (OR, 95CI 13.5,
6.1-29.8 13,5.1-33 6.9, 2.8-17.1) compared to
the N-AHS. Conclusion - These data suggest that
swine influenza viral infections are common among
swine workers and their non-exposed-to-swine
spouses. Agriculture workers should be given
special attention in influenza surveillance
strategies and in planning for influenza
pandemics. Abstract revised
1Center for Emerging Infectious Diseases, Dept.
of Epidemiology, University of Iowa College of
Public Health, Iowa City, IA 2Department of
Epidemiology, University of Iowa College of
Public Health, Iowa City, IA 3Department of
Pathobiological Sciences, University of
Wisconsin-Madison, Madison, WI 4Strain
Surveillance Section, Influenza Branch, Centers
for Disease Control and Prevention, Atlanta, GA
5National Cancer Institute, Bethesda, MD
6Immunology and Viral Pathogenesis Section,
Centers for Disease Control and Prevention,
Atlanta, GA
To validate serological outcomes, culture and
molecular studies will be performed on gargle and
swab specimens to identify influenza and specific
hemagglutinin types.

• Human-to-swine and swine-to-human influenza
  transmission have been well-documented.
• Olsen et. al. found that 23 of 74 swine farm
  workers had serologic evidence of emergent swine
  influenza infection as compared to controls (Emerg Infect Dis 20028814-9).
• Myers et. al. found farmers, meat processing
  workers, and veterinarian workers all to have
  higher odds of elevated antibodies against swine
  H1N1 and swine H1N2 influenza viruses compared to
  nonexposed controls (Clin Infect
  Dis,20064214-20).

• Many of the non-exposed subjects were spouses of
  the exposed, and although they reported little
  swine exposure, they had lived on a swine farm
  (Table 2).
• Odds for having elevated antibodies against swine
  influenza viruses were high for the exposed and
  their non-exposed spouses compared to University
  of Iowa non-exposed controls (Table 3).

• These data suggest that swine workers are at
  increased risk of swine influenza virus
  infections.
• Swine worker spouses may also be at increased
  risk of infection through indirect swine viruses.
  
• Agriculture workers should be given special
  attention in influenza surveillance strategies
  and in planning for influenza pandemics.

• We used data from the 90,000-person Agricultural
  Health Study to identify 707 study subjects who
  were occupationally exposed to swine and/or
  poultry, and a gender and age-group matched group
  of 80 control subjects without such exposures.
• Upon enrollment, and after 12 and 24 months of
  follow-up, participants are asked to complete a
  questionnaire and to donate sera (589 AHS-E and
  69 AHS-NE).
• During the 2 years of follow-up, if a participant
  develops signs and symptoms of an influenza-like
  illness, they complete a questionnaire, provide a
  gargle sample and a nasal swab within 48 hours of
  symptom development, and ship the questionnaire
  and gargle sample to the investigators.
• Use hemagglutination inhibition and
  microneutralization procedures to examine sera.
• Hypotheses testing will be performed for any
  swine (H1, H3) or avian (H3, H4, H5, H6, H7, H9)
  influenza hemagglutinin type and later stratified
  by each swine and avian type.
• Serologic results will be adjusted for antibodies
  to human influenza virus.

• 803 rural Iowans enrolled from 29 counties
• Enrollment characteristics compared to nonexposed
  University of Iowa control group (Table 1)

Coinvestigators - Troy McCarthy, Ana W. Capuano,
MPS, Charles F. Lynch, MD, PhD, Debbie A.
Wellman, Kelly A. Lesher, Sharon F. Setterquist,
MT(ASCP), Norma J. Miller, Patricia A P Gillette,
MPH, Christopher W. Olsen, DVM, PhD, Alexander I.
Klimov, PhD, Michael C. Alavanja, PhD, and Jackie
M. Katz, PhD Collaborators - Kevin Knudson,
PhDCEID Staff - Whitney Baker, Mark Lebeck,
Ghazi Kayali Funding NIH / NIAID- R21
AI059214-01

• 721 (90 ) participated in the 12-month
  follow-up.
• Thus far 64 influenza-like-illness specimens have
  been received, and 16 influenza A isolates have
  been recovered 15 human H3N2 and 1 swine H1N?
  virus.
• After only 12 months of follow-up, the exposed
  have evidence of an increase in antibody titer
  against both swine influenza viruses that is not
  explained by antibodies to human H1 viruses
  (Table 4).