The Mighty Vector: Role of vector ecology in virulence ...

0 downloads 0 Views 1MB Size Report
Isabel Rodríguez-‐Barraquer, Sunil S. Solomon, Periaswamy Kuganantham, ... Pachamuthu Balakrishnan, Suniti Solomon, Shruti H. Mehta, Derek A. T.
The  Mighty  Vector:  Role  of  vector  ecology  in   virulence  evolution.     Nitesh  Vinodbhai  Pandey  ,     Researcher,  Indian  Astrobiology  Research  Centre,  Mumbai,  India.   Address  correspondence  to  Nitesh  V.  Pandey,  [email protected]    

 

Abstract:     Ewald  has  suggested  that  the  vector-­‐borne  diseases  should  evolve  to  be  more   virulent  than  the  diseases  transmitted  directly.    I  have  argued  in  this  paper  that   the  fundamental  assumptions  of  Ewald's  model  regarding  the  vector-­‐borne   diseases  are  flawed  and  are  also  the  root  cause  of  its  discrepancies.    I  have   refined  Ewald's  trade-­‐off  model  of  Virulence  evolution  by  incorporating  the   insights  from  vector  ecology.  The  vector-­‐parasite  system  of  Aedes  aegypti  and   dengue  has  been  selected  to  demonstrate  the  strong  role  of  vector  ecology  as  a   major  determinant  of  virulence  evolution.  I  have  also  shown  that  the  Darwinian   fitness  of  parasite  variants,  be  it  mild  or  the  virulent  one,  would  depend  on   ecological  aspects  of  the  vector.  A  thought  experiment  has  also  been  proposed  to   illustrate  this  improvised  model  of  virulence  evolution.     Keywords:  Trade-­‐off  model;  Virulence  evolution;  Vector-­‐borne  diseases;  Dengue   epidemiology  

    1.  Introduction:     What  according  to  you  is  the  most  important  thing  that  should  regularly  happen   in  Science?  If  you  had  to  take  my  opinion,  I  would  have  certainly  gone  with   “Debates”.  Yes,  Debates  must  regularly  take  place  in  science  as  they  are  at  the   very  core  of  its  foundation  and  are  required  for  its  progress.  Nobody  within  the   Evolutionary  Biology  fraternity  would  have  ever  imagined  that  a  fellow   researcher's  (Ewald)  fateful  bout  of  diarrhea  would  usher  a  productive  debate   regarding  virulence  evolution  of  infectious  diseases  [1].  Ewald  asked  some   interesting  questions  about  the  nature  of  his  diarrhea  symptoms,  which  he  got   while  working  on  a  research  project  near  garbage  dump  on  the  outskirts  of   Manhattan,  Kansas  [1].  This  experience  inspired  him  to  think  about  the  disease   from  the  pathogen’s  perspective.  The  application  of  evolutionary  insights  he   derived  from  this  incidence  helped  him  to  formulate  his  trade-­‐off  model  of   virulence  evolution  [1].  Ewald  proposed  that  the  difference  in  severity  of  the   infectious  diseases  could  be  understood  from  their  different  modes  of   transmission  [1].  He  suggested  that  parasites  often  have  variants,  which  differ  in   terms  of  their  virulence.  The  Darwinian  fitness  of  these  pathogenic  variants  

would  depend  on  the  modes  through  which  they  are  transmitted  from  an   infected  host  to  the  susceptible  one  [1].  Ewald  suggested  that  for  a  disease,   which  is  transmitted  directly,  the  virulent  pathogenic  variant  that  can  make  the   human  host  too  sick  or  bedridden  would  have  least  chances  of  reaching  to  a  new   susceptible  host.  This  meant  that  directly  transmitted  diseases  would  evolve  to   be  mild  since  the  milder  variants  would  out-­‐compete  the  virulent  variants  in   terms  of  transmission.  They  would  also  have  higher  Darwinian  fitness.  However,   if  there  are  vectors  to  spread  the  virulent  variants  of  diseases  from  one  host  to   another,  the  pathogen  can  afford  to  be  deadly  and  exploitative  as  the  cost  paid   for  the  resulting  severe  sickness  is  less.  In  short,  the  vector-­‐borne  diseases   would  evolve  to  be  highly  virulent  because  they  have  to  pay  lesser  cost  for  the   diminished  mobility  of  the  host  that  results  from  their  excessive  exploitation.  In   spite  of  its  intuitive  appeal  this  model  has  been  criticized  on  various  grounds  so   far.  The  following  paragraph  deals  with  the  exact  issues  of  Ewald’s  model  in   detail.        

2.  The  problem:  Contrary  evidence  for  the  majority  of  the   vector-­‐borne  diseases.    

In  contrary  to  what  Ewald  has  predicted,  most  of  the  vector-­‐borne  diseases  are   either  mild  or  asymptomatic  in  the  majority  of  the  infected  cases.  There  are  in   fact  diseases  like  River  blindness  [2],  Zika  [3]  and  Chikungunya  [4],  which  in   spite  of  being  vector-­‐borne  rarely  kill  their  host.  Even  diseases  like  Dengue  [5]   and  Malaria  [6]  cause  severe  symptoms  in  a  very  small  portion  of  the  total   infected  population.  The  extreme  cases  that  we  observe  in  such  diseases  are   nothing  but  the  tip  of  an  iceberg.       Ewald's  model  and  the  overall  trade-­‐off  theory  of  virulence  evolution  have  been   widely  challenged  since  last  30  years  mainly  for  its  lack  of  evidence.  Most  of   these  authors  who  have  challenged  the  model  have  pointed  towards  various   shortcomings  to  justify  the  observed  discrepancies.  Their  respective  arguments   are  as  follows:   1.  Trade-­‐off  model  is  over  simplistic  [7].   2.  The  biological  details  of  parasite's  life  history  and  disease  etiology  were  never   considered  within  the  model  [7].   3.  The  role  of  immune  response  in  causing  disease  was  overlooked  [8].   4.  There  is  ambiguity  over  the  definition  of  virulence.  Morbidities  like  Anemia,   Weight  loss  and  infertility  should  also  be  considered  as  a  measure  of  virulence   apart  from  host  death  [7].   5.  There  might  be  no  correlation  between  excessive  replication  of  the  parasite   and  virulence.  A  trade-­‐off  between  Virulence  and  transmission  might  be   completely  absent  in  some  vector-­‐borne  diseases  [8].     For  a  vector-­‐borne  disease  with  no  correlation  between  higher  replication  rates   of  parasites  and  virulence,  it  is  justified  that  such  diseases  can  be  mild.  However,   How  can  we  justify  the  mild  and  asymptomatic  nature  of  Dengue  and  Malaria  in   the  majority  of  the  infected  cases?  Although  we  have  a  clear  evidence  that  a  

virulence-­‐transmission  trade-­‐off  exists  for  both  of  these  diseases  [9,10].  As  per   Ewald's  model,  Dengue  and  Malaria  should  have  been  highly  virulent  in  the   majority  of  the  cases.  This  theory  and  empiricism  mismatch  for  the  vector-­‐borne   diseases  have  led  some  researchers  to  doubt  the  potential  of  the  trade-­‐off  model   [7].  This  is  certainly  not  an  encouraging  situation  and  demands  a  resolution  for   sure.  In  the  following  section,  I  have  argued  that  the  fundamental  assumptions  of   Ewald's  model  for  the  vector-­‐borne  diseases  are  unnecessary  and  flawed.  These   assumptions  are  also  the  root  cause  of  discrepancies  in  his  model.  I  have  also   offered  a  resolution  that  will  make  his  model  more  robust  and  would  also   address  the  major  discrepancies.    

    3.  The  Solution:  Incorporating  insights  from  Vector-­‐ecology  to   make  Ewald's  model  more  robust.  

  I  would  like  to  use  the  case  of  the  Dengue  vector  Aedes  aegypti  to  prove  my   following  claims  i.e      1.  Ewald's  assumptions  for  the  vector-­‐borne  diseases  are   flawed      2.  Insights  from  Vector-­‐ecology  can  improvise  his  trade-­‐off  model  of   virulence  evolution.     In  the  following  Section,  I  will  be  presenting  my  arguments  against  the   fundamental  assumptions  of  Ewald  regarding  virulence  evolution  among  the   vector-­‐  borne  diseases.     3.1.    Flawed  Assumption  1:  Mobility  of  the  host  is  not  important  for  the   transmission  of  vector-­‐borne  diseases  because  vectors  can  transmit  the   virulent  variants  of  the  parasite  from  the  immobile  host  to  susceptible  one   [1].     For  the  above  assumption  to  hold  true  it  would  require  that  the  vector  is  highly   mobile  so  that  it  could  travel  long  distances  to  spread  the  disease  from  a   bedridden  severely  sick  host  to  a  new  uninfected  host.  However,  here  comes  an   important  twist.  The  Aedes  aegypti  vector  that  transmits  Dengue  is  very  lazy  and   sessile.  This  vector  has  a  mean  dispersal  rate  of  50m  and  maximum  dispersal  of   100m  [11].  It  stays  in  only  one  home  for  the  entire  lifespan.  The  vector  density  is   in  fact  so  low  that  there  is  sometime  just  one  Aedes  aegypti  mosquito  for  the   entire  house  [12].  The  infection  rate  of  Aedes  aegypti  even  in  the  endemic  areas   is  as  low  as  1%  [13].    In  such  cases,  a  vector-­‐borne  disease  can  spread  or   perpetuate  only  if  the  host  is  mobile  enough  to  infect  the  new  vectors  so  that  the   transmission  cycle  could  continue.  Therefore  a  vector-­‐borne  disease  that  causes   acute  illness  and  limits  the  mobility  of  host  would  easily  go  extinct  if  both  the   vector  and  the  host  have  restricted  mobility.  This  has  serious  implications  for   virulence  evolution  among  those  vector-­‐borne  diseases  for  which  there  exists  a   trade-­‐off  between  virulence  and  transmission.  I  would  like  to  explain  this  claim   of  mine  from  the  following  thought  experiment:    

Imagine  that  there  are  two  best  friends,  Sam  and  John.  Sam  is  infected  with  a   mild  variant  of  Dengue  virus  whereas  John  is  infected  with  the  virulent  variant  of   the  same  virus.  Since  Sam  is  infected  with  a  mild  variant  he  will  be  asymptomatic   and  mobile  but  John  who  is  infected  with  the  virulent  variant  will  be  bedridden   and  severely  sick.  The  severely  sick  host  John  has  a  probability  of  being  bitten  by   the  mosquitoes  of  his  own  house  only,  whereas  the  mobile  host  Sam  has  a   probability  of  being  bitten  by  mosquitoes  of  every  home  he  visits,  be  it  his   friend’s  or  relative’s.  Sam  can  also  infect  many  Aedes  aegypti  mosquitoes  at     cluster  regions  like  Shops  and  markets  where  Aedes  aegypti  density  might  be  bit   higher.  This  means  that  the  mild  variants  of  Dengue  virus  would  be  transmitted   far  more  than  the  one,  which  are  virulent.  The  Darwinian  fitness  of  variants   causing  mild  disease  would  be  much  higher  over  the  virulent  ones.  Therefore  a   stable  evolutionary  strategy  for  any  such  vector-­‐borne  parasite  would  be  to   cause  a  mild  infection  because  it  significantly  enhances  their  transmission   benefits.  This  would  eventually  lead  to  decrease  in  virulence  of  the  disease  with   time  in  contrary  to  what  Ewald  had  predicted  in  his  model.     The  following  is  a  visual  representation  of  my  thought  experiment:     Figure  1:  Thought  Experiment  on  the  role  on  vector  ecology  in  virulence   evolution  in  dengue.    

 

      3.2.  Flawed  Assumption  2:  The  higher  virulence  of  the  vector-­‐borne   parasite  is  adaptive  as  it  makes  the  reservoir  host  less  defensive  towards   vector  bite  and  therefore  facilitates  transmission  [1].     The  explanation  offered  for  assumption  1  equally  proves  the  fallacy  of  this   assumption  too.  However,  it  is  important  to  trace  the  roots  of  this  assumption.   Ewald  had  derived  this  assumption  from  a  study  done  on  mice,  which  concluded   that  mice  with  a  higher  magnitude  of  malaria  parasites  were  more  susceptible  to   mosquito  bites  compared  to  the  mice  that  were  diseases  free  [1].  Similar  animal   studies  for  Dengue  were  also  cited  to  support  this  assumption  [9].  The   conclusion  of  these  animal  studies  that  more  aggressively  replicating  strains   incapacitate  the  host  and  would  have  higher  chances  of  being  transmitted  does   not  apply  to  the  real  world  scenario.  In  reality,  the  virulent  strains  of  Dengue   cripple  the  mobility  of  the  host  and  hinder  the  process  of  transmission.  These   laboratory  models  of  the  vector-­‐borne  diseases  hardly  capture  the  real  world   complexity  of  human  host,  mosquito  vector  and  parasite  interactions  and   therefore  the  generalization  derived  from  such  studies  have  resulted  in   discrepancies.     I  would  also  like  to  challenge  Ewald’s  view  that  higher  virulence  is  adaptive  for   vector-­‐borne  parasites  because  it  facilitates  transmission.  The  following  is  my   critique:  We  usually  defend  ourselves  from  something  we  know  would  do  us   harm  or  would  attack  us  in  some  way.  What  if  we  are  attacked  in  a  way  that  we   do  not  even  realize?  Aedes  vector  employ  this  exact  strategy  to  bite  their  host.   The  Aedes  aegypti  vector  mainly  bites  at  the  ankle,  which  is  one  of  the  safest   regions  to  bite  [14].  The  biting  of  the  mosquito  also  often  goes  unnoticed  [14].  I   do  not  think  that  any  change  in  behavior  that  makes  the  host  less  defensive   towards  the  mosquito  bite  is  required  because  mosquitoes  already  have   adaptations  that  serve  the  same  purpose.  I  would  like  to  make  similar  arguments   for  the  primary  Malaria  vectors  i.e  Anopheles  gambiae  and  Anopheles  funestus.   Both  of  these  vectors  have  their  peak  biting  time  in  the  midnight  when  the   human  host  is  in  deep  sleep  and  inactive  [15].  Be  it  sleeping  human  host  or  the   sick  one  lying  on  the  bed,  both  are  equally  incapable  of  defending  themselves   from  the  biting  mosquitoes.  The  malaria  vectors  like  the  ones  of  dengue  are  also   known  to  bite  at  the  ankles  [15].       In  the  following  section,  I  would  like  to  show  the  utility  of  applying  the  insights   from  Vector-­‐ecology  to  Ewald’s  trade-­‐off  model  of  virulence  evolution.        

   

4.  This  new  model  that  integrates  the  ecology  of  the  vector  with   Ewald's  conventional  trade-­‐off  model  has  two  natural   predictions  to  offer:  

  1.  The  mobility  of  the  host  would  be  very  important  in  the  transmission   dynamics  of  dengue.     2.  Majority  of  the  dengue  cases  would  be  asymptomatic  in  the  host.  However,   these  asymptomatic  hosts  would  have  enough  viremia  to  infect  the  biting   mosquito.     I  would  like  to  present  pieces  of  evidence  available  from  dengue  endemic  regions   to  support  the  above  two  predictions  of  this  new  model.     4.1.    The  role  of  human  movement  in  transmission  dynamics  of  Dengue:     Recent  evidence  from  the  field  studies  of  dengue  endemic  regions  has  concluded   that  in  the  absence  of  movement  of  the  human  host  the  dengue  transmission   cycle  would  collapse  [16].  The  regular  infection  of  new  Aedes  aegypti  mosquito   by  the  viremic  human  host  is  what  drives  dengue  epidemic  and  not  vice  versa.   Results  from  a  recent  study  on  Dengue  epidemiology  also  indicated  that  the   human  movement  alone  could  explain  significant  spatial  variation  in  urban   transmission  rates  of  Dengue  [16].     A  study  in  Thailand  that  used  DNA  fingerprinting  to  track  the  feeding  behaviour   of  Aedes  mosquitoes  concluded  that  frequent  and  heterogeneous  biting  by  Aedes   aegypti  on  residents  and  transient  visitors  and  mosquito  feeding/transmission   hotspots  are  important  entomologic  features  of  dengue  epidemiology  [17,18].   Padmanabha  et.  al  also  concluded  that  human  social  behaviour  and  demography   drive  patterns  of  fine-­‐scale  dengue  transmission  in  endemic  areas  of  Colombia   [19].  A  Cambodia  based  study  also  suggested  that  human  movements  could  only   explain  the  heterogeneous  propagation  of  dengue  infection  [20].     Studies  done  in  dengue-­‐endemic  countries  like  Thailand  and  Peru  have   concluded  that  the  fine-­‐scale  movement  of  the  reservoir  hosts  is  the  major  driver   of  dengue  [21].  These  studies  also  implied  that  the  geographical  spread  of  the   dengue  is  not  possible  without  the  movement  of  dengue  carriers  [21].  The   overall  engagement  of  people  in  their  regular  and  daily  routine  determines  their   exposure  to  the  insect  vectors  and  plays  a  very  important  role  in  the  dynamics  of   pathogen  transmission  [21].  Stoddard  et.  al  based  on  their  activity  space  model   concluded  that  dengue  is  mostly  transmitted  when  people  are  mobile  as  well  as  are   engaged  in  their  daily  activities.  Reiner  et.al  have  shown  that  it  is  the  friends  and   relatives  of  the  infected  dengue  subject  who  have  much  higher  chances  of  getting   the  secondary  infection  and  not  their  neighbours  [22].  This  argument  is  derived   from  a  simple  logic  that  the  most  probable  places  that  any  infected  but  mobile   host  will  visit  are  its  social  contacts.  This  has  been  rightly  named  as  the  Social   contact  hypothesis  proposed  to  explain  the  Dengue  epidemiology  [22].    

Stoddard  et  al.  in  their  brilliant  paper  have  emphasized  the  role  of  human   movement  as  a  major  driver  not  only  for  dengue  but  also  for  most  of  the  vector-­‐ borne  diseases  [21].  The  patterns  of  the  human  movement  are  the  most  important   determinant  of  vector  exposure  and  therefore  they  must  be  considered  in  every   study  done  to  understand  epidemiology  of  vector-­‐borne  diseases  [21].  

      4.2.    The  asymptomatic  hosts  are  capable  of  infecting  mosquitoes  as  well  as   they  constitute  the  majority  of  the  dengue  cases:     Three-­‐quarters  of  the  estimated  390  million  dengue  virus  (DENV)  infections   each  year  are  clinically  in  apparent  [23].  People  with  asymptomatic  dengue  virus   infections  were  generally  considered  dead-­‐end  hosts  for  transmission  because  it   was  assumed  that  viremia  in  such  carriers  would  be  too  low  to  infect  the  biting   mosquito.  The  recent  field  studies  have  presented  a  different  picture  though.   Despite  of  their  lower  average  level  of  viremia,  people  with  inapparent  infections   can  be  infectious  to  mosquitoes  [24].  Moreover,  at  a  given  level  of  viremia,  a   dengue  virus  infected  person  with  no  detectable  symptoms  or  before  the  onset  of   symptoms  is  significantly  more  infectious  to  mosquitoes  than  people  with   symptomatic  infections  [24].  This  is  because  people  infected  with  dengue   without  clinical  symptoms  may  be  exposed  to  more  mosquitoes  through  their   undisrupted  daily  routines  than  people  who  are  severely  sick.  The  asymptomatic   hosts  represent  the  bulk  of  dengue  positive  infections;  data  collected  from  recent   dengue  studies  indicate  that  the  asymptomatic  reservoir  hosts  might  be   significantly  contributing  to  regular  virus  infections  of  mosquitoes  than   previously  recognized  [24].         A  variety  of  the  cohort  studies  done  to  understand  the  transmission  pattern  of   dengue  in  endemic  countries  like  Thailand  [25-­‐29],  Indonesia  [30,31],  Nicaragua   [32],  Vietnam  [33],  Peru  [34,35]  and  India  [36]  have  supported  the  strong  role  of   asymptomatic  cases  in  the  spread  of  dengue.  These  studies  have  also  concluded   that  the  transmission  dynamics  of  dengue  can  be  heterogeneous  in  nature  both   temporally  as  well  as  spatially.  Studies  that  have  incorporated  geographical   cluster  designs  to  understand  dengue  transmission  also  support  the  role  of   asymptomatic  dengue  carriers  in  its  transmission  to  a  substantial  level.   [37,38,39]     A  recent  study  in  a  Leishmaniasis  endemic  region  also  found  that  the   asymptomatic  cases  are  far  more  prevalent  than  the  symptomatic  cases  [40].  A   future  study  in  the  Visceral  Leishmaniasis  endemic  region  could  confirm  the  role   of  asymptomatic  carriers  in  the  transmission  cycle  of  this  disease  too.  A    study   done  to  understand  Malaria  epidemiology  confirmed  that  the  asymptomatic   carriers  not  only  are  in  significant  majority  but  they  also  infect  the  Anopheles   mosquito.  The  asymptomatic  malaria  carriers  therefore  contribute  in  the   transmission  of  the  protozoal  disease  in  a  major  proportion  [6].    

4.3.    Dengue  Mathematical  Models  to  support  the  importance  of  host   mobility  in  Dengue  transmission:     Apart  from  studies  collecting  data  from  dengue  endemic  regions  there  are   simulations  and  mathematical  models  that  have  emphasized  the  role  of  host   movement  as  a  mandatory  requirement  for  dengue  transmission.  [41,42].  One  of   the  studies  that  analyzed  the  trade-­‐off  between  host  immune  response  and  levels   of  arbovirus  viremia  suggested  the  following  relation:  the  strains  of  arbovirus   that  can  keep  ‘low  profile’  in  the  vertebrate  host  and  maintain  viremia  for  a   longer  period  of  time  would  have  maximum  transmission  benefits  and  therefore   also  Darwinian  fitness.  The  study  also  implied  that  the  strains  producing  the   higher  magnitude  of  viremia  get  cleared  by  the  immune  system  quickly  and   rarely  get  transmitted  [43].  These  results  have  some  important  implications  for   the  virulence  evolution  in  case  of  dengue  as  well  as  for  other  vector  borne   diseases  where  the  mobility  of  the  reservoir  host  is  equally  important  for   transmission.     4.4.    The  problem  with  the  evidence  of  transmission-­‐virulence  trade-­‐off  in   dengue:     There  have  been  many  studies  that  have  confirmed  a  trade-­‐off  between   transmission  and  virulence  for  dengue  infections.  The  following  are  some   important  and  common  conclusions  of  these  studies:   1. Dengue  virus  variants,  which  reproduce  aggressively  and  lead  to  higher   viremia  are  more  virulent.  There  have  been  multiple  independent  studies   that  have  confirmed  a  correlation  between  higher  viremia  titer  with   dengue  severity  [9].     2. Laboratory  studies  on  dengue  serotype  2  have  found  that  genotypes  that   are  more  virulent  to  humans  also  more  readily  infect  mosquitoes  and   human  cells.  Elevated  human  viremia  is  then  supposed  to  increase  the   chance  of  transmission  to  mosquitoes  [9].     3. It  has  been  suggested  that  people  harbouring  more  virulent  parasites  may   be  easier  prey  for  mosquitoes,  thereby  increasing  the  frequency  of  being   bitten.  Some  authors  have  also  suggested  that  the  biting  rate  increases   with  virulence  to  humans  [9].       All  these  studies  that  have  correlated  virulent  strains  and  their  higher  infectivity   for  mosquitoes  have  one  common  drawback.  These  studies  were  performed  in   artificial  laboratory  settings  and  were  based  on  animal  models.    Data  from   artificial  blood  meals  suggest  that  arboviruses  would  have  to  reach  quite  high   titers  in  vertebrate  hosts  before  significant  transmission  could  occur.  However,   this  inference  may  not  be  true  because  the  threshold  for  transmission  of  a  given   arbovirus  is  substantially  lower  when  a  vector  feeds  on  a  living  vertebrate  host  

rather  than  an  artificial  meal  [43].  In  fact,  the  same  authors  have  also  pointed  out   that  in  case  of  higher  viremia  the  overall  infectivity  of  the  host  for  the  biting   mosquito  may  decrease  because  of  the  host  immune  response  [43].  Also,  none  of   these  studies  considered  the  important  ecological  aspects  of  Aedes  aegypti  like   its  limited  dispersal,  very  low  density  and  low  infectivity.  This  is  the  reason  we   do  not  find  proof  for  Ewald’s  prediction  in  case  of  vector-­‐borne  diseases.  In   reality  what  we  observe  is  that  the  mild  strains  are  more  infectious  to  the   mosquitoes.  They  also  get  more  bitten  by  vectors  because  the  infected  host  is   mobile  and  can  visit  new  places  having  their  own  vectors.  The  insights  from   vector  ecology  when  integrated  with  Ewald’s  trade-­‐off  model  offers  a  picture   that  is  much  more  realistic  epidemiologically.        

5.  Can  insights  from  Vector  ecology  of  Aedes  aegypti  also  explain   the  epidemiology  of  Chikungunya?  

  The  rationale  of  this  refined  trade-­‐off  model  also  applies  to  other  diseases   transmitted  by  Aedes  aegypti  mosquitoes  like  Yellow  fever,  Chikungunya  and   Zika.  Yellow  fever  and  Zika  are  asymptomatic  or  cause  mild  illness  in  the   majority  of  the  infected  individuals.  Chikungunya  unlike  dengue  and  Zika  causes   symptoms  like  high  fever  and  severe  joint  pain  in  more  than  80%  of  the  cases   and  therefore  severely  cripples  mobility  of  the  host.  This  new  improvised  model   suggests  that  Chikungunya  would  always  be  a  clustered  disease  confined  to   smaller  geographical  regions.  This  is  because  the  disease  has  a  negative  effect  on   host  mobility,  which  we  know  is  important  for  transmission  [44].    This  model   therefore  also  explains  why  Chikungunya  occurs  in  small  clusters  of  a  limited   region  of  300  meters  unlike  wider  local  spread  of  dengue.  The  outbreaks  of   Chikungunya  are  also  sporadic.  A  study  done  to  understand  the  transmission   dynamics  and  distribution  pattern  of  Chikungunya  outbreak  in  a  village  of   Dhaka,  Bangladesh  came  at  the  following  conclusions  that  perfectly  fit  with  this   refined  trade-­‐off  model  of  Ewald  [44].  The  following  were  the  conclusions  of  this   study:     1.        The  risk  of  getting  infected  with  Chikungunya  significantly  dropped  for   people  living  50m  away  from  the  household,  which  had  positive  cases  [44].     2.        Females  were  1.5  times  more  likely  to  become  infected  than  males,  which   was  virtually  identical  to  the  relative  risk  of  being  at  home  estimated  from  an   independent  human  movement  study  in  the  country  [44].     I  would  like  to  predict  that  the  local  spread  of  Chikungunya  would  never  be  as   intense  as  dengue  because  of  its  severe  symptoms  that  hamper  the  mobility  of   the  host.  In  future,  if  we  find  Chikungunya  to  spread  in  the  endemic  locations   with  similar  intensity  to  that  of  dengue  then  it  will  mean  that  there  are  a  lot  of   asymptomatic  cases  of  Chikungunya  too  and  they  also  go  undetected.  Based  on   the  insights  of  this  improved  trade-­‐off  model  of  Ewald  we  can  categorize  any   vector-­‐borne  diseases  in  terms  of  their  potential  to  quickly  spread  within  

geographical  regions.  This  would  be  a  very  powerful  tool  as  it  might  help  us  in   predicting  as  well  as  controlling  important  vector-­‐borne  diseases.        

6.  A  general  refined  trade-­‐off  model  of  Virulence  Evolution:   Integrating  Vector  ecology  with  Ewald’s  model.    

The  disease  vectors  are  highly  diverse  in  various  aspects  like  their  flight  range,   density,  lifespan,  vectorial  capacity  and  biting  time.  These  different  facets  of   vector  would  affect  the  virulence  evolution  of  every  concerned  vector-­‐parasite   system  differently.  However,  the  most  important  determinants  of  virulence   evolution  are  vector’s  flight  range  and  contact  rate  with  the  host.  The  following   table  covers  the  diversity  among  the  vectors  of  the  most  important  human   diseases.     Table  1:  Disease  vectors  and  their  diversity.    

 

 

        It  is  clear  from  the  case  of  dengue  that  the  insights  from  vector-­‐ecology  can  make   Ewald’s  trade-­‐off  model  much  better.  The  question  that  naturally  arises  is  that  if   using  the  similar  insights  from  other  disease-­‐vector  systems,  could  we  also  build   a  general  model  that  offers  testable  predictions  for  them  too?  The  answer  would   be  yes.  The  following  are  some  general  predictions  on  the  virulence  evolution  of   vector-­‐borne  diseases  based  on  this  improvised  model.  The  model  can  predict   about  the  relative  Darwinian  fitness  of  the  milder  and  the  virulent  variants  for  a   disease  having  a  trade-­‐off  between  virulence  and  transmission.  The  predictions   are  as  follows:     1.  The  Darwinian  fitness  of  the  mild  strains  of  the  parasites  would  be   higher  for  a  disease  in  which  mobility  of  the  host  is  important  in   transmission  dynamics.  The  mobility  of  the  host  is  crucial  for  a  disease  that  is   transmitted  by  a  vector  with  the  following  features:  a.  very  limited  dispersal,  b.   low  infectivity  and  c.  low  density.  Ex.  Aedes  aegypti  and  Sandfly.  In  case  of  a   disease  transmitted  by  a  sessile  vector  with  very  limited  dispersal  and   infectivity,  it  is  the  mild  parasite  strains  that  will  have  higher  Darwinian  fitness,   which  would  also  easily  outnumber  the  virulent  strains  that  make  the  reservoir   host  immobile  because  of  incapacitation.     The  following  figure  captures  some  very  important  ecological  aspects  of   Dengue  vector  Aedes  aegypti:     Figure  2:  Important  ecological  features  of  Aedes  aegypti    

 

      2.  In  certain  special  cases  mobility  of  the  host  might  also  be  required  even   if  the  vector  can  travel  or  disperse  for  longer  distances.  The  following  are   the  traits  of  vector  that  would  also  lead  to  higher  Darwinian  fitness  of  mild   parasitic  strains:  Vector  has  not  only  low  density  and  infectivity  but  it  also  has  a   very  low  contact  frequency  with  the  reservoir  host.  In  such  cases  since  the   contact  of  the  vector  and  host  are  very  limited  the  perpetuation  of  the   transmission  cycle  would  sustain  only  under  any  of  the  two  conditions.    1.  Vector   has  a  very  long  lifespan  of  at  least  some  months  or  a  year  2.  The  reservoir  host   suffers  from  a  chronic  infection  and  has  enough  parasites  in  the  blood  to  infect   the  vectors  for  years  or  at  least  months.  This  would  ensure  that  the  new  vectors   are  regularly  infected  to  keep  the  transmission  cycle  going.  Ex.  Tsetse  fly  that   transmits  the  trypanosomes  to  humans  and  animals  to  cause  Human  African   trypanosomiases,  commonly  called  as  Sleeping  Sickness  [45].  The  tsetse  fly  has  a   longer  flight  range  and  dispersal  rate  but  its  frequency  of  contact  with  the   human  hosts  is  still  very  less.  Therefore  long  flight  range  is  not  much  of  use  in   disease  transmission  here  [46,47].  To  make  the  situation  worse,  the  salivary   gland  infection  rates  in  the  Tsetse  fly  have  been  found  to  rarely  exceed  0.1%   even  in  the  endemic  regions  [46,47].  The  tsetse  fly  bites  the  human  host  during   the  daytime  when  people  are  accidentally  exposed  to  the  diseases  vectors  while   performing  their  regular  duties  in  the  riverine  tsetse  belts.  The  tsetse  flies  are   also  strongly  attracted  towards  moving  object  [48].  Humans  travelling  on   vehicles  are  therefore  found  to  be  more  susceptible  to  tsetse  bites  [48].  This  also   implies  that  immobile,  sick  and  bedridden  host  is  complete  dead  end  case  and   would  never  contribute  to  the  transmission  chain  of  sleeping  sickness.   Additionally,  terminal  stage  of  Sleeping  sickness,  which  involves  neurological   issues,  leads  to  complete  immobility  of  the  host,  which  is  again  detrimental  to   the  transmission  cycle  of  the  disease.  For  a  vector-­‐borne  disease  in  which  the   vector-­‐host  contact  frequency  is  very  low  and  the  vector  itself  has  low  infectivity   of  as  less  as  0.1%,  the  mobility  of  the  host  becomes  even  more  important.  A  very   natural  question  that  must  be  answered  is  that  how  does  the  sleeping  sickness   causing  pathogen  perpetuates  given  the  fact  that  it  is  intrinsically  virulent  and  it   also  cripples  the  host  mobility  in  the  terminal  stages?  As  per  the  insights  of   vector  ecology  such  pathogen  can  only  survive  by  causing  a  chronic  infection  so   that  the  reservoir  host  has  transmissible  parasites  density  in  the  blood  for  a   broader  time  period.  This  strategy  would  ensure  that  the  reservoir  hosts   contribute  towards  infecting  new  vectors  and  help  the  perpetuation  of  the   disease.  The  pathogen  therefore  by  causing  chronic  infection  would  also  incur   less  cost  from  the  death  of  the  host,  as  the  host  would  have  already  contributed   towards  the  transmission  cycle  before  his  or  her  death.  A  mobile  host  with   chronic  transmissible  parasitaemia  is  required  for  a  vector  that  has  very  low   infectivity,  low  density  and  low  frequency  of  interaction  with  the  host.       This  improvised  model  also  challenges  one  of  the  Ewald’s  idea  that  vector-­‐borne   disease  circulating  more  in  the  human  host  would  become  virulent  with  time.   The  majority  of  the  sleeping  sickness  in  humans  is  caused  by  Trypanosoma   brucei  gambiense  whereas  Trypanosoma  brucei  rhodesiense  mainly  infects  

animals  [45].  Trypanosoma  brucei  rhodesiense,  which  has  circulated  less  in  the   human  population,  kills  the  human  host  in  months  whereas  the  more  prevalent   and  circulated  Trypanosoma  brucei  gambiense  takes  years  to  do  the  same  [45].   Based  on  this  model  I  would  like  to  argue  that  the  virulent  variant  of  sleeping   sickness  parasite  that  is  transmitted  by  a  vector  of  very  low  infectivity  and  low   contact  frequency  couldn’t  afford  to  kill  the  host  early.  The  early  death  of  the   human  host  would  drastically  reduce  the  probabilities  of  new  vector  infections.   This  would  also  lead  to  collapse  of  the  transmission  cycle  or  least  transmission   benefits  for  the  parasite.  Parasitic  variants  that  keep  the  human  host  alive  for  a   longer  period  of  time  would  certainly  out-­‐compete  the  ones  that  kill  the  host  too   early  in  this  particular  case.       The  frequency  of  interaction  between  the  vector  and  the  reservoir  host  in  case  of   River  blindness  or  Onchocerciasis  is  also  very  less  [49].       This  refined  trade-­‐off  model  is  equally  applicable  for  all  those  diseases,  which   are  transmitted  by  sessile  vectors  with  very  limited  dispersal  like  Aedes  aegypti.   The  Sand  fly  vector  responsible  for  transmission  of  Visceral  leishmaniasis  are   weak  fliers,  travelling  with  a  characteristic  short  hopping  flight,  and  usually,   disperse  not  more  than  a  few  hundred  metres  from  their  breeding  sites  [21].   Leishmaniasis  very  much  like  Dengue  is  asymptomatic  in  a  majority  of  the  cases   [40].  However,  there  are  no  studies  done  as  of  yet  to  prove  the  role  of   asymptomatic  cases  in  the  transmission  of  visceral  leishmaniasis.  Mathematical   modelling  suggests  that  these  asymptomatic  carriers  constitute  a  reservoir  of   parasites  driving  the  epidemic  [40],  although  their  infectiousness  to  sand  flies  is   not  yet  formally  established.     Based  on  this  new  refined  trade-­‐off  model  of  virulence  evolution  I  would  like  to   hypothesize  that  the  asymptomatic  cases  of  leishmaniasis  would  turn  out  to  be   the  major  contributor  in  its  transmission  dynamics  too.  I  wish  that  in  future   experimental  studies  would  be  undertaken  to  prove  the  role  of  asymptomatic   subjects  in  the  transmission  dynamics  of  visceral  leishmaniasis.     This  model  can  also  be  used  for  diseases  that  have  a  trade-­‐off  between   transmission  and  virulence  and  are  transmitted  by  the  Culex  mosquitoes.  West   Nile  Virus  is  one  of  the  arboviruses  that  are  transmitted  by  Culex  mosquitoes   where  this  refined  trade-­‐off  model  can  be  used  to  come  up  with  testable   predictions  [50,51].  The  model  can  be  also  used  to  understand  the  epidemiology   of  lymphatic  filariasis  also  transmitted  by  Culex  Pipiens  [52].     .      

     

7.  The  most  important  refutable  prediction  of  this  new   improvised  model  of  Ewald:    

The  asymptomatic  cases  of  most  of  the  vector-­‐borne  diseases  should  play  a  far   more  important  role  in  the  transmission  dynamics  of  the  diseases  compared  to   what  is  known  as  of  now.  I  would  suggest  that  the  researchers  involved  in   epidemiological  studies  of  vector-­‐borne  diseases  should  also  investigate  and   therefore  vindicate  the  role  of  asymptomatic  cases  in  overall  epidemiology  of  the   disease.  I,  however,  believe  that  for  a  disease  transmitted  by  Vectors  having   limited  dispersal,  low  infectivity  and  low  frequency  of  contact  with  the  host,  the   role  of  asymptomatic  humans  in  the  transmission  of  the  disease  is  going  to  be   significantly  higher.        

8.  The  limitations  of  this  improvised  Ewalds’s  trade-­‐off  model:  

  There  have  been  multiple  studies  since  2009  that  have  confirmed  the  role  of   reservoir  host  movement  in  the  transmission  of  dengue.  Similar  results  from   other  regions  in  the  world  where  dengue  is  prevalent  might  help  to  add  more   weightage  to  my  claims.  Though  the  evidence  cited  and  available  is  in  no  way  in   sufficient.     Virulence,  as  pointed  out  by  Bull  and  Ebert,  is  a  complex  trait  and  there  are   multiple  parameters  involved  into  it  [53].  In  case  of  Dengue,  the  immune  system   of  the  host  plays  a  significant  role  in  the  extreme  but  rare  cases  of  Dengue   Haemorrhagic  fever  and  Dengue  Shock  Syndrome.  The  insights  from  this  paper   are  based  on  the  verified  assumption  that  variants  that  reproduce  more   vigorously  lead  to  more  severe  immune  response  and  therefore  lead  to  severe   symptomatic  Dengue  cases  that  result  in  incapacitation  of  the  host.       I  have  not  considered  the  case  of  secondary  infections  of  the  same  host  with  a   different  dengue  strains.  The  secondary  dengue  infections  that  lead  to  dengue   shock  syndrome  and  dengue  hemorrhagic  fever  are  rare.  The  secondary   infections  would  not  affect  the  predictions  of  this  model  in  a  major  way.      

9.  Conclusions:     I  strongly  believe  in  this  famous  quote  of  Albert  Einstein  i.e  “The  important  thing   is  not  to  stop  questioning”  [54].  Sometimes  a  theory  or  a  scientific  model  can  be   improved  by  questioning  its  fundamental  assumptions.  I  have  shown  clearly  in   this  paper  that  correcting  the  flawed  assumptions  in  Ewald’s  model  actually   results  in  a  more  robust  framework.  This  new  improved  model  that  integrates   vector  ecology  into  Ewald’s  conventional  model  of  virulence  evolution  has  much   to  offer  towards  understanding  of  infectious  disease  epidemiology.  The  model   not  only  offers  new  predictions  but  also  answers  many  of  the  puzzling  questions   that  conventional  model  was  not  able  to  explain.  I  would  like  to  suggest  that  

further  field  studies  should  focus  on  the  relevant  role  of  vector  ecology  in   virulence  evolution  of  other  diseases  too.  Researchers  involved  in  mathematical   modelling  of  infectious  diseases  could  improve  their  models  by  including   variables  related  to  vector  ecology.  In  short,  there  lies  a  highly  fertile  ground  for   the  researchers  right  at  the  intersection  of  vector  ecology,  host  movement  and   virulence  evolution.  Time  has  served  us  with  a  wonderful  opportunity  and  we   must  all  must  jump  into  it  with  full  conviction.      

  Acknowledgements:  

  I  would  like  to  acknowledge  the  sincere  support  of  Dr.  Stephen  Stearns  for  his   suggestions  on  improvement  of  this  manuscript.  I  would  also  like  to   acknowledge  the  constant  support  I  received  from  my  mentor  Pushkar  Ganesh   Vaidya.    

Conflict  of  Interest:  

  The  author  declares  that  he  does  not  have  any  conflict  of  interest.    

References:      

1. Ewald  PW,  Evolution  of  infectious  diseases,  Oxford,  UK:  Oxford  University   Press,  1994.     2. Progress  towards  eliminiating  onchocerciasis  in  the  WHO  region  of  the   Americas:  Verification  of  elimination  of  transmission  in  Guatemala  and   progress  report  on  the  elimation  of  human  onchocerciasis,  2015-­‐2016".     World  Health  Organization.  March  2017.  Retrieved  17  March  2017     3. Saiz  J-­‐C,  Martín-­‐Acebes  MA,  Bueno-­‐Marí  R,  et  al,    Zika  Virus:  What  Have  We   Learnt  Since  the  Start  of  the  Recent  Epidemic?  Frontiers  in  Microbiology,  8   (2017),  1554.  doi:10.3389/fmicb.2017.01554.     4. Weaver  SC,  Osorio  JE,  Livengood  JA,  Chen  R,  Stinchcomb  DT,  Chikungunya   virus  and  prospects  for  a  vaccine,  Expert  review  of  vaccines,  11  (2012),   1087-­‐1101.  doi:10.1586/erv.12.84     5. Duong  V,  Lambrechts  L,  Paul  RE,  Ly  S,  Lay  RS,  Long  KC,  et  al,   Asymptomatic  humans  transmit  dengue  virus  to  mosquitoes,  Proc  Natl  Acad   Sci  USA,  112  (2015),  14688– 93.  https://doi.org/10.1073/pnas.1508114112  PMID:26553981    

6. Mallika  Imwong,  Kasia  Stepniewska,  Rupam  Tripura,  Thomas  J.  Peto,  Khin   Maung  Lwin,  Benchawan  Vihokhern,  Klanarong  Wongsaen,  Lorenz  von   Seidlein,  Mehul  Dhorda,  Georges  Snounou,  Lilly  Keereecharoen,  Pratap   Singhasivanon,  Pasathorn  Sirithiranont,  Jem  Chalk,  Chea  Nguon,  Nicholas   P.  J.  Day,  Francois  Nosten,  Arjen  Dondorp,  Nicholas  J.  White;  Numerical   Distributions  of  Parasite  Densities  During  Asymptomatic  Malaria,  The   Journal  of  Infectious  Diseases,  213  (2016),  1322–1329,   https://doi.org/10.1093/infdis/jiv596     7. ALIZON,  S.,  HURFORD,  A.,  MIDEO,  N.  and  VAN  BAALEN,  M,  Virulence   evolution  and  the  trade-­‐off  hypothesis:  history,  current  state  of  affairs  and   the  future,  Journal  of  Evolutionary  Biology,  22  (2009),  245–259.   doi:10.1111/j.1420-­‐9101.2008.01658.x     8. Day  T,  Graham  AL,  Read  AF,  Evolution  of  parasite  virulence  when  host   responses  cause  disease,  Proceedings  of  the  Royal  Society  B:  Biological   Sciences,  274  (2007),  2685-­‐2692.  doi:10.1098/rspb.2007.0809.    

 

9. Medlock  J,  Luz  PM,  Struchiner  CJ,  Galvani  AP,  The  Impact  of  Transgenic   Mosquitoes  on  Dengue  Virulence  to  Humans  and  Mosquitoes,  The  American   naturalist,  174  (2009),  565-­‐577.  doi:10.1086/605403.     10. Mackinnon  MJ,  Read  AF,  Virulence  in  malaria:  an  evolutionary  viewpoint,   Philosophical  Transactions  of  the  Royal  Society  B:  Biological  Sciences,   359(2004),965-­‐986.  doi:10.1098/rstb.2003.1414.     11. Getis  A,  Morrison  A,  Gray  K,  Scott  T,  Characteristics  of  the  spatial  pattern   of  the  dengue  vector,  Aedes  aegypti,  in  Iquitos,  Peru,  Am  J  Trop  Med  Hyg  69   (2003),  494–505.       12. Schneider  JR,  Morrison  AC,  Astete  H,  Scott  TW,  Wilson  ML  (2004),  Adult   size  and  distribution  of  Aedes  aegypti  (Diptera:  Culicidae)  associated  with   larval  habitats  in  Iquitos,  Peru.  J  Med  Entomol  41  (2004),  634–642.         13. De  Benedictis  J,  Chow-­‐Shaffer  E,  Costero  A,  Clark  GG,  Edman  JD,  et  al,   Identification  of  the  people  from  whom  engorged  Aedes  aegypti  took  blood   meals  in  Florida,  Puerto  Rico,  using  polymerase  chain  reaction-­‐based  DNA   profiling,  Am  J  Trop  Med  Hyg  68  (2003),  437–446.      

  14. https://www.cdc.gov/dengue/resources/30Jan2012/aegyptifactsheet.pd f    

15. Braack  L,  Hunt  R,  Koekemoer  LL,  et  al,  Biting  behaviour  of  African  malaria   vectors:  1.  where  do  the  main  vector  species  bite  on  the  human   body?  Parasites  &  Vectors,  8  (2015).  doi:10.1186/s13071-­‐015-­‐0677-­‐9.    

16. Stoddard  ST,  Forshey  BM,  Morrison  AC,  et  al,  House-­‐to-­‐house  human   movement  drives  dengue  virus  transmission,  Proceedings  of  the  National   Academy  of  Sciences  of  the  United  States  of  America,  110(2013),  994-­‐999.   doi:10.1073/pnas.1213349110.     17. Harrington  LC,  Fleisher  A,  Ruiz-­‐Moreno  D,  Vermeylen  F,  Wa  CV,  et  al,   Heterogeneous  Feeding  Patterns  of  the  Dengue  Vector,  Aedes  aegypti,  on   Individual  Human  Hosts  in  Rural  Thailand,  PLoS  Negl  Trop  Dis,  8(2014),   e3048.  doi:10.1371/journal.pntd.0003048       18. Yoon  I-­‐K,  Getis  A,  Aldstadt  J,  et  al,  Fine  Scale  Spatiotemporal  Clustering  of   Dengue  Virus  Transmission  in  Children  and  Aedes  aegypti  in  Rural  Thai   Villages.  Barrera  R,  ed,  PLoS  Neglected  Tropical  Diseases,  6  (2012),  e1730.   doi:10.1371/journal.pntd.0001730  

  19. Padmanabha  H,  Correa  F,  Rubio  C,  Baeza  A,  Osorio  S,  Mendez  J,  et  al,   Human  Social  Behavior  and  Demography  Drive  Patterns  of  Fine-­‐Scale   Dengue  Transmission  in  Endemic  Areas  of  Colombia,  PLoS  ONE,  10  (2015),   e0144451.  https://doi.org/10.1371/journal.pone.0144451       20. Teurlai  M,  Huy  R,  Cazelles  B,  Duboz  R,  Baehr  C,  Vong  S,  Can  Human   Movements  Explain  Heterogeneous  Propagation  of  Dengue  Fever  in   Cambodia?,  PLoS  Negl  Trop  Dis6  (2012),  e1957.   https://doi.org/10.1371/journal.pntd.0001957     21. Stoddard  ST,  Morrison  AC,  Vazquez-­‐Prokopec  GM,  Paz  Soldan  V,  Kochel   TJ,  et  al,  The  Role  of  Human  Movement  in  the  Transmission  of  Vector-­‐  Borne   Pathogens,  PLoS  Negl  Trop  Dis  3(2009),  e481.   doi:10.1371/journal.pntd.0000481    

 

22. Reiner  RC,  Stoddard  ST,  Scott  TW,  Socially  structured  human  movement   shapes  dengue  transmission  despite  the  diffusive  effect  of  mosquito   dispersal,  Epidemics,  6  (2014),  30-­‐36.  doi:10.1016/j.epidem.2013.12.003.   23. Bhatt  S,  et  al,  The  global  distribution  and  burden  of  dengue,  Nature,  496   (2013),  504–507     24. Duong  V,  Lambrechts  L,  Paul  RE,  et  al,  Asymptomatic  humans  transmit   dengue  virus  to  mosquitoes,  Proceedings  of  the  National  Academy  of  

 

 

 

Sciences  of  the  United  States  of  America,  112  (2015),  14688-­‐14693.   doi:10.1073/pnas.1508114112.     25. Sangkawibha  N,    Rojanasuphot  S,    Ahandrik  S,  et  al,  Risk  factors  in  dengue   shock  syndrome:  a  prospective  epidemiologic  study  in  Rayong,  Thailand.  I.   The  1980  outbreak,  Am  J  Epidemiol  ,    120  (1984),  653-­‐69.     26. Burke  DS,    Nisalak  A,    Johnson  DE,    Scott  RM,  A  prospective  study  of  dengue   infections  in  Bangkok,  Am  J  Trop  Med  Hyg  ,  38  (1988),  172-­‐80.     27. Endy  TP,    Chunsuttiwat  S,    Nisalak  A,  et  al.  Epidemiology  of  inapparent  and   symptomatic  acute  dengue  virus  infection:  a  prospective  study  of  primary   school  children  in  Kamphaeng  Phet,  Thailand,  Am  J  Epidemiol  ,  156   (2002),  40-­‐51.     28. Endy  TP,    Nisalak  A,    Chunsuttiwat  S,  et  al,  Spatial  and  temporal  circulation   of  dengue  virus  serotypes:  a  prospective  study  of  primary  school  children  in   Kamphaeng  Phet,  Thailand,  Am  J  Epidemiol  ,  156  (2002),  52-­‐59.     29. Mammen  MP,    Pimgate  C,    Koenraadt  CJ,  et  al,  Spatial  and  temporal   clustering  of  dengue  virus  transmission  in  Thai  villages,  PLoS  Med  ,  5   (2008),  pg.  e205       30. Graham  RR,    Juffrie  M,    Tan  R,  et  al,  A  prospective  seroepidemiologic  study   on  dengue  in  children  four  to  nine  years  of  age  in  Yogyakarta,  Indonesia  I.   studies  in  1995–1996,  Am  J  Trop  Med  Hyg  ,  61  (1999),  412-­‐19.  

 

 

  31. Porter  KR,    Beckett  CG,    Kosasih  H,  et  al,  Epidemiology  of  dengue  and   dengue  hemorrhagic  fever  in  a  cohort  of  adults  living  in  Bandung,  West   Java,  Indonesia,  Am  J  Trop  Med  Hyg  ,  72  (2005),  60-­‐6.     32. Balmaseda  A,    Standish  K,    Mercado  JC,  et  al,  Trends  in  patterns  of  dengue   transmission  over  4  years  in  a  pediatric  cohort  study  in  Nicaragua,  J  Infect   Dis  ,  201  (2010),  5-­‐14.     33. Tien  NT,    Luxemburger  C,    Toan  NT,  et  al,  A  prospective  cohort  study  of   dengue  infection  in  school  children  in  Long  Xuyen,  Viet  Nam,  Trans  R  Soc   Trop  Med  Hyg  ,  104  (2010),  592-­‐600.     34. Rocha  C,    Morrison  AC,    Forshey  BM,  et  al,  Comparison  of  two  active   surveillance  programs  for  the  detection  of  clinical  dengue  cases  in  Iquitos,   Peru,  Am  J  Trop  Med  Hyg  ,  80  (2009),  656-­‐60.  

   

 

 

35. Morrison  AC,    Minnick  SL,    Rocha  C,  et  al,  Epidemiology  of  dengue  virus  in   Iquitos,  Peru  1999  to  2005:  interepidemic  and  epidemic  patterns  of   transmission,  PLoS  Negl  Trop  Dis  ,  4  (2010),  e670  .     36. Isabel  Rodríguez-­‐Barraquer,  Sunil  S.  Solomon,  Periaswamy  Kuganantham,   Aylur  Kailasom  Srikrishnan,  Canjeevaram  K.  Vasudevan,  Syed  H.  Iqbal,   Pachamuthu  Balakrishnan,  Suniti  Solomon,  Shruti  H.  Mehta,  Derek  A.  T.   Cummings,The  Hidden  Burden  of  Dengue  and  Chikungunya  in  Chennai,   India,  PLOS  Neglected  Tropical  Diseases,  9  (2015),  e0003906  DOI:   10.1371/journal.pntd.0003906     37. Endy  TP,    Anderson  KB,    Nisalak  A,  et  al,  Determinants  of  inapparent  and   symptomatic  dengue  infection  in  a  prospective  study  of  primary  school   children  in  Kamphaeng  Phet,  Thailand,  PLoS  Negl  Trop  Dis  ,  5   (2011),  e975  .     38. Beckett  CG,    Kosasih  H,    Faisal  I,  et  al,  Early  detection  of  dengue  infections   using  cluster  sampling  around  index  cases,  Am  J  Trop  Med  Hyg  ,  72   (2005),  777-­‐82.     39. Reyes  M,    Mercado  JC,    Standish  K,  et  al,  Index  cluster  study  of  dengue  virus   infection  in  Nicaragua,  Am  J  Trop  Med  Hyg  ,  83  (2010),  683-­‐9.       40. Singh  OP,  Hasker  E,  Sacks  D,  Boelaert  M,  Sundar  S,  Asymptomatic   Leishmania  Infection:  A  New  Challenge  for  Leishmania  Control.  Clinical   Infectious  Diseases:  An  Official  Publication  of  the  Infectious  Diseases   Society  of  America,  58  (2014),  1424-­‐1429,  doi:10.1093/cid/ciu102     41. Liebman  KA,  Stoddard  ST,  Morrison  AC,  Rocha  C,  Minnick  S,  et  al,  Spatial   Dimensions  of  Dengue  Virus  Transmission  across  Interepidemic  and   Epidemic  Periods  in  Iquitos,  Peru  (1999–2003),  PLoS  Negl  Trop  Dis  6   (2012),  e1472.  doi:10.1371/journal.pntd.0001472     42. Adams  B,  Kapan  DD,  Mosquito:  Understanding  the  Contribution  of  Human   Movement  to  Vector-­‐Borne  Disease  Dynamics,  PLoS  ONE  4  (2009),  e6763.   https://doi.org/10.1371/journal.pone.0006763  

  43. Tortoise  or  the  Hare  Approach:  Althouse  BM,  Hanley  KA,  The  tortoise  or   the  hare?  Impacts  of  within-­‐host  dynamics  on  transmission  success  of   arthropod-­‐borne  viruses,  Philosophical  Transactions  of  the  Royal  Society   B:  Biological  Sciences,  370  (2015),  20140299.   doi:10.1098/rstb.2014.0299    

44. Salje  H  et  al,  How  social  structures,  space  and  behaviors  shape  the  spread  of   infectious  diseases  using  chikungunya  as  a  case  study,  Proc.  Natl  Acad.  Sci.   USA,  113  (2016),  420–25,  (doi:10.1073/pnas.1611391113)  

  45. Franco  JR,  Simarro  PP,  Diarra  A,  Jannin  JG,  Epidemiology  of  human  African   trypanosomiasis,  Clinical  Epidemiology,  6  (2014),  257-­‐275.   doi:10.2147/CLEP.S39728.     46. Service  M,  Medical  Entomology  for  Students,  5th  ed.  Cambridge:   Cambridge  University  Press;   2012.  https://doi.org/10.1017/CBO9781139002967.     47. Msangi  AR,  Whitaker  CJ,  Lehane  MJ,  Factors  influencing  the  prevalence  of   trypanosome  infection  of  Glossina  pallidipes  on  the  Ruvu  flood  plain  of   Eastern  Tanzania,  Acta  Trop,  70  (1998),  143-­‐55.  10.1016/S0001-­‐ 706X(98)00013-­‐8.       48. Rayaisse  J-­‐B,  Salou  E,  Courtin  F,  et  al,  Baited-­‐boats:  an  innovative  way  to   control  riverine  tsetse,  vectors  of  sleeping  sickness  in  West  Africa,  Parasites   &  Vectors,  8  (2015),  236,  doi:10.1186/s13071-­‐015-­‐0851-­‐0.     49. Winthrop  KL,  Furtado  JM,  Silva  JC,  Resnikoff  S,  Lansingh  VC,  River   Blindness:  An  Old  Disease  on  the  Brink  of  Elimination  and  Control,  Journal   of  Global  Infectious  Diseases,  3  (2011),  151-­‐155.  doi:10.4103/0974-­‐ 777X.81692.    

 

50. Ciota,  Alexander  T.  et  al,  Dispersal  of  Culex  Mosquitoes  (Diptera:  Culicidae)   From  a  Wastewater  Treatment  Facility,  Journal  of  Medical  Entomology,   49.1  (2012):  35–42.     51. Andreadis  S,  Dimotsiou  O,  Savopoulou-­‐Soultani  M,  Variation  in  adult   longevity  of  Culex  pipiens  f.  pipiens,  vector  of  the  West  Nile  Virus,  Parasitol   Res,  113(2014),  4315–9.  

 

52. Hairston  NG,  De  Meillon  B,  On  the  inefficiency  of  transmission   of  Wuchereria  bancrofti  from  mosquito  to  human  host,  Bull  World  Health   Organ,  38  (1968),  935–941.  

    53. Stearns  SC,  Koella  Jk  ,  Evolution  in  Health  and  Disease,  Second   Edition  (Oxford  University  Press,  Oxford),  2008.    

54. W.  Berger,  A  More  Beautiful  Question:  The  Power  of  Inquiry  to  Spark   Breakthrough  Ideas,  Bloomsbury  Publishing,  USA,  2014.