Risk of exposure to Lyme disease in northern California

1
Risk of exposure to Lyme disease in northern
California
Caryl Waggett, Environmental Science, Allegheny
College and Robert Lane, Insect Biology,
University of California at Berkeley

• Lyme disease
• Most prevalent vector-borne disease in the
  United States. Human incidence in California is
  patchy, reflecting underlying variability in
  vector density and infection prevalence
• Three-part study designed to identify high risk
  habitats, develop monitoring program and conduct
  epidemiological assessment of risk factors
• Ecological assessment
• Four dominant habitats (n48) and ecotones
  (n10) flagged for ticks
• Collected and analyzed gt3,500 ticks bi-weekly
  over 3 years
• Acarolgic risk, the number of infected ticks/100
  m2, varied by
• Habitat, season, tick life stage
• Hardwood forests pose greatest risk of exposure
  to infected nymphs
  Ecotones and chaparral pose moderate risk
  to infected adults
• Abiotic factors correlated with acarologic
  risk included
• Aspect, slope, canopy, leaf duff, substrate, pH,
  relative humidity and temperature

Mean number of infected ticks per 100 m2
The Lyme disease spirochete, Bb, has a
complicated life cycle involving two vectors and
many hosts that affect risk of human exposure.
The tick responsible for transmitting Bb to
humans has definitive habitat preferences and
seasonal distributions.
Lyme disease is the most prevalent in the
Northeast and Midwest, but high numbers of human
cases have been identified in coastal northern
California.
Month

• Model results for monitoring Lyme disease risk
• High risk habitats not uniquely identifiable
  using satellite imagery
• Overall spectral signatures appear unique, but
  high band variability present within each habitat
• TM5 Model Accurate Kappa statistic
  100
• Overall model (n351) 72 accurate 66
• Water (n90) 100 accurate 100
• Open grassland (n56) 92 accurate 91
• Hardwood forests (n67) 61 accurate 53
• Chaparral (n54) 52 accurate 42
• Woodland grass (n84) 48 accurate 37
• Tasseled cap model had similar accuracy
  (K100)64
• Mixed hardwood forests have highest acarologic
  risk. Inability of model to accurately
  characterize this habitat suggests that TM5
  imagery may not be sufficiently sensitive for
  long-term and rapid monitoring of human risk of
  Lyme disease in northern California
• Models from Northeast and Midwest with higher
  accuracy characterized extreme habitats (e.g.,
  forests v. meadows, or wooded lots v. lawns) as
  opposed to two similar forest types
• Hyperspectral or higher resolution imagery may
  be necessary to adequately characterize
  different forested habitats in region

• Monitoring Lyme disease risks
• Rapid changes in land use patterns in rural
  California
• CA population to double from 1990-2020 to 60
  million people
• Rural northern CA currently affordable,
  accessible
• Increased habitat fragmentation, development in
  high risk habitats
• Model developed using satellite imagery to
  characterize high risk habitats identified from
  ecological assessments
• Landsat TM5 imagery (30m resolution) from Nov
  2002
• Drought- and winter-deciduous trees had lost
  leaves
• 89 reference sites (1,792 pixels) from 5
  habitats
• Hardwood forests, woodland grass, grass,
  chaparral, control water
• Values from 6 non-thermal bands used to create
  habitat-specific reference signatures
• Each of 51 validation sites (351 pixels) were
  compared to reference signatures
• Each pixel characterized to habitat using root
  mean square (RMS) error with smallest difference
• Model re-run using Tasseled cap indices
• Greenness, wetness, brightness

Band 4
Band 4

• Epidemiological assessment of risk factors
• Cross-sectional survey of community at high risk
  for Lyme disease, 14-year follow-up of
  serological survey and questionnaire conducted in
  1988-89 (Lane et al. 1992)
• 43 cumulative prevalence of Lyme disease, 1.5
  annual incidence
• 11 of 76 variables examined significantly
  associated with exposure to Lyme disease
  pathogen, Borrelia burgdorferi (Bb)
• Length of residence, annual time at residence,
  time spent outdoors, seasonal use of property,
  habitat diversity, property management,
  gardening, caring for animals, ownership of farm
  animals, stargazing, and lack of protective
  clothing
• Activities, behaviors placed residents at high
  risk for many tick-borne diseases
• Lyme disease, human granulocytic and monocytic
  ehrlichiosis, babesiosis, bartonellosis
• Despite rural environment and behavioral
  patterns (e.g., time spent outdoors), the
  majority of risk factors were peri-domestic in
  naturesimilar to Northeast and Midwest
• Educational efforts conducted 14 years earlier
  have proved effective

All residents spend an extraordinary amount of
time outdoors, but residents who have been
exposed to Bb spend statistically more time
outdoors than non-exposed residents. Residents
spend less time outdoors now than they did in
1988-89.
Residents in rural northern California live
adjacent to habitats at high risk from infected
nymphs and along ecotones at moderate risk from
infected adult ticks.
These studies were supported by generous grants
from NIH and CDC to R.S.L.