physical activity and health risks

25
1 Physical Activity and Health Risks Meghan Nairn Research completed in partial fulfillment of the requirements For the degree of Master of Science in Demography Center for Demography & Population Health Florida State University August 2011 COMMITTEE APPROVAL ________________________________________ Professor Isaac W. Eberstein, Chair ________________________________________ Professor Elwood D. Carlson

Upload: meghan-nairn

Post on 22-Aug-2015

40 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: Physical Activity and Health Risks

  1  

 

Physical  Activity  and  Health  Risks    

Meghan  Nairn  Research  completed  in  partial  fulfillment  of  the  requirements  

For  the  degree  of  Master  of  Science  in  Demography  

Center  for  Demography  &  Population  Health  Florida  State  University  

August  2011                        

COMMITTEE  APPROVAL        

________________________________________  Professor  Isaac  W.  Eberstein,  Chair  

   

________________________________________  Professor  Elwood  D.  Carlson  

 

   

   

Page 2: Physical Activity and Health Risks

  2  

   

 ABSTRACT  

  This  paper  will  examine  the  relationship  between  level  of  physical  activity  

and  health  related  behaviors.  The  purpose  is  to  examine  if  less  active  or  more  active  

people  are  more  or  less  likely  to  see  doctors  for  an  annual  medical  check  –up,  if  

more  active  people  are  less  likely  to  smoke,  and  also  if  more  active  people  are  less  

likely  to  drink.  Previous  literature  will  be  used  as  a  background  for  this  study  and  to  

explain  some  connections  that  have  already  been  made  between  these  variables.  

Demographics  including  age,  sex,  race,  income  and  education  will  be  controlled.  Self-­‐

reported  health  will  also  be  used  to  control  for  health  status.  The  data  used  were  

taken  from  BRFSS  2007  telephone  survey,  which  asks  health  related  questions  to  

those  18  and  older.  Logistic  regression  was  used  to  find  that  as  activity  levels  

increase  ,  one  is  less  likely  to  see  a  doctor  within  the  year,  less  likely  to  smoke,  and  

more  likely  to  drink.  These  relationships  change  directions  when  demographics  are  

controlled  for  and  change  within  each  model  when  self-­‐reported  health,  age,  sex,  

income,  and  education  are  controlled  for.  It  was  concluded  that  the  least  active  

group  has  higher  odds  than  the  highest  active  group  to  have  seen  a  doctor  for  a  

medical  check-­‐up  within  the  past  year,  those  with  moderate  activity  level  have  the  

highest  odds  of  smoking,  and  those  with  the  highest  activity  level  have  the  highest  

odds  of  being  a  moderate  or  heavy  drinker  compared  to  being  a  non-­‐drinker.    

 

 

 

 

 

 

 

 

 

Page 3: Physical Activity and Health Risks

  3  

INTRODUCTION  

  The  purpose  of  this  study  is  to  examine  the  relationship  between  physical  

activity  and  health  behavior.  The  objective  is  to  explore  the  association  between  a  

person’s  level  of  physical  activity  and  risky  health  behaviors  such  as  smoking  and  

drinking.  It  will  also  test  if  activity  is  associated  with  annual  doctor  visits  by  asking  if  

a  person  has  seen  a  physician  within  the  past  year.  Although  the  direction  of  the  

relationship  is  ambiguous  to  predict,  the  expected  result  is  that  there  will  be  an  

association  between  activity  level  and  doctor  visits  and  those  who  are  more  active  

will  smoke  and  drink  less  than  those  who  are  less  active.    

 

PREVIOUS  LITERATURE  

  Many  studies  have  noted  the  positive  health  effects  of  physical  activity.  It  has  

been  linked  to  reduced  premature  mortality  in  numerous  ways.  Physical  activity  can  

reduce  the  risks  of  coronary  heart  disease,  hypertension,  colorectal  cancer,  obesity,  

and  osteoporosis  (  Blair,1996).  Physical  activity  can  help  reduce  the  risk  of  

overweight  or  obesity,  which  leads  to  decline  in  physical  health.  One  study  showed  

that  even  if  physical  activity  did  not  lead  to  extreme  weight  loss,  for  those  

individuals  who  are  obese,  physical  activity  may  still  provide  health  benefits  

(Penedo  &  Dahn,  2005).  Physical  activity  can  also  have  positive  effects  on  mental  

health.  It  can  improve  confidence,  well-­‐being,  anxiety  reduction  and  intellectual  

functioning.  (U.S.  Department  of  Health  &  Human  Services,  1996).    

 

  Some  authors  have  concluded  that  a  person’s  involvement  in  physical  activity  

can  be  influenced  by  various  factors.  Self  efficacy  has  been  shown  to  be  a  correlate  

of  physical  activity  (  Trost  et  al.  2002).  Demographics  have  also  been  noted.  Men  

have  been  observed  to  have  higher  levels  of  physical  activity  than  women,  Whites  

are  more  physically  active  than  Blacks  and  Hispanics,  education  and  income  are  

both  positively  associated  with  physical  activity,  and  participation  in  physical  

activity  declines  with  age  (Casperan  et  al.  1992,  Casperan  et  al.  1995).  Some  studies  

have  also  reported  that  physical  activity  can  have  an  impact  on  a  person’s  mental  

health  in  addition  to  their  physical  health.  Physical  activity  can  improve  confidence,  

Page 4: Physical Activity and Health Risks

  4  

well-­‐being,  anxiety  reduction  and  intellectual  functioning  (Hughes  1984).  McAuley  

1994  has  shown  that  there  are  correlations  between  exercise  and  self-­‐steem,  self-­‐

efficacy,  psychological  well-­‐being  and  cognitive  function  as  well  as  anxiety,  stress  

and  depression.  

  There  are  different  ways  to  look  at  the  relationship  between  physical  activity  

and  health  behavior.  Some  examine  physical  activity  as  having  an  impact  on  health,  

while  others  examine  a  person’s  health  as  having  an  impact  on  their  level  of  activity.  

Warburton  et  al.  2006  reviewed  the  health  benefits  of  physical  activity  from  the  

view  point  that  physical  activity  is  what  impacts  health.  They  concluded  that  

physical  activity  does  contribute  to  the  prevention  of  some  chronic  diseases  and  

reduces  the  risk  of  premature  death.  Increased  activity  in  both  men  and  women  

reduces  the  relative  risk  of  cardiovascular  related  death,  and  increased  energy  

expenditure  was  associated  with  a  mortality  benefit  of  20%.  Increased  energy  

expenditure  also  decreased  the  incidence  of  type  2  diabetes  by  6%,  and  weight  loss  

through  diet  and  exercise  reduced  the  incidence  of  high  risk  individuals  by  40%-­‐

60%  over  a  4  year  period.  Physical  activity  can  also  impact  certain  cancers,  

specifically  colon  and  breast.  Men  and  women  showed  30%-­‐40%  reduction  in  the  

relative  risk  of  colon  cancer,  and  physically  active  women  showed  20%-­‐30%  

reduced  risk  of  breast  cancer  compared  to  their  inactive  counterparts  (U.S.  

Department  of  Health  and  human  Services  1996).  

  The  other  direction  to  be  considered  is  that  a  person’s  existing  health  

condition  will  affect  their  ability  to  engage  in  physical  activity.  The  Center  for  

Disease  Control  reports  that  many  older  adults  live  with  ,  rather  than  die  from,  

disabling  chronic  disease.  Chronic  disease  or  physical  handicaps  may  reduce  a  

person’s  mobility  which  would  lead  to  a  sedentary  lifestyle.  Attempts  to  change  

levels  of  activity  are  met  by  many  barriers.  One  study  found  that  older  adults  are  

less  likely  than  younger  adults  to  engage  in  leisure-­‐time  physical  activity  when  they  

perceive  their  neighborhood  as  unsafe  (  Center  for  Disease  Control,1999).  Poor  

perceived  health,  pain  or  fear  of  pain  as  a  result  of  a  chronic  disease  may  also  

influence  a  sedentary  lifestyle  compared  to  a  physically  active  one  (  Brawley,  2003).    

Page 5: Physical Activity and Health Risks

  5  

  Cigarette  smoking  is  noted  as  the  top  preventable  cause  of  mortality  in  the  

United  States.    It  is  estimated  that  8.6  million  American  suffer  from  conditions  

caused  by  cigarette  smoking  such  as  selected  cancers,  heart  attack,  stroke,  chronic  

bronchitis  and  emphysema  (Center  for  Disease  Control  2002;  2003).  Since  reports  

like  these  have  come  out  over  the  years  the  rates  of  smoking  have  started  to  decline,  

however  smoking  habits  are  not  the  same  for  all  demographic  groups  and  have  

shown  to  vary  across  subgroups  of  race,  ethnicity,  age,  and  social  class  (Keife  2001,  

Peirce  1989,Fiore  1989,Flint  1998).  

  According  to  the  Centers  for  Diseases  Control,  as  of  2010  an  estimated  46  

million  adults  smoke  cigarettes  in  the  United  States.    Current  smokers  are  defined  as  

those  who  have  smoked  at  least  100  cigarettes  in  their  lifetime  and  currently  smoke  

cigarettes  some  days  or  every  day.  More  men  than  women  fall  into  this  category  

with  23.5%  of  adult  males  being  current  smokers  and  17.9%  of  adult  females  being  

smokers.  Twenty-­‐two  percent  of  whites  smoke,  closely  followed  by  21%  of  blacks  

and  14.5%  of  Hispanics.  Those  with  lower  education  had  higher  percentages  of  

smoking.  Of  those  who  have  a  High  School  diploma,  49%  are  smokers,  33.6  %  of  

adults  with  9  to  11  years  of  education  are  smokers,  11.1%  for  adults  with  an  

undergraduate  degree  and  5.6%  of  adults  who  have  a  graduate  college  degree.  In  

terms  of  life  expectancy  of  smokers,  those  who  quit  smoking,  on  average,  live  longer  

than  those  who  are  continuous  smokers.  Those  who  have  never  smoked  have  the  

longest  life  expectancies,  followed  by  those  who  quit  smoking,  with  continuous  

smokers  having  the  shortest  life  expectancy  (  Nam  et  al  1994,  Rogers  and  Powell-­‐

Grinner  1991).  

   

  Close  behind  smoking  among  preventable  causes  of  death  is  alcohol  abuse,  

which  is  ranked  third  in  the  United  States  (McGinnis  1993).  (Second  is  high-­‐blood  

pressure,  which  has  smoking,  diet,  and  exercise  as  major  risk  factors[  Danaei  et  al.,  

2009]).  Binge  drinking  relates  to  alcohol  related  deaths  in  multiple  ways  and  is  

generally  defined  as  5  or  more  alcoholic  beverages  on  1  occasion  (Weshcler  1998,  

Weschler  2001).  One  study  focused  on  the  sociodemographic  characteristics  of  

older  adults  and  unhealthy  drinking  patterns.  It  was  found  that  women  are  more  

Page 6: Physical Activity and Health Risks

  6  

likely  than  men  to  report  no  drinking  and  less  likely  to  report  unhealthy  

consumption,  heavier  alcohol  consumption  is  more  prevalent  among  Whites  than  

any  other  race,  and  higher  levels  of  education  and  income  are  associated  with  a  

higher  prevalence  of  drinking  (Merrick  2006).  

  Previous  research  has  shown  that  95%  of  women  regardless  of  age  or  race  

seek  medical  care  within  a  year’s  time.  The  most  frequently  cited  reasons  for  visits  

were  medical  examination  followed  by  progress  visits.  This  is  33%  higher  than  

males’  rates  of  seeking  ambulatory  medical  care  within  the  past  year.  The  number  of  

visits  increases  as  age  increases.  The  percentage  of  Blacks  who  had  an  ambulatory  

visit  within  the  past  year  was  slightly  higher  than  the  percentage  of  whites.  Black  

women  were  three  times  more  likely  than  white  women  to  be  covered  by  Medicaid,  

which  covered  9%  of  visits  (Brett  2001).  

  This  research  will  examine  the  relationship  between  level  of  physical  activity  

and  health  related  behaviors.  The  behaviors  will  be  medical  exam  within  the  past  12  

months,  smoking,  and  being  a  moderate  or  heavy  drinker  versus  a  non-­‐drinker.  

Variables  are  constructed  by  using  definitions  set  by  Centers  for  Disease  Control  as  

well  as  other  studies  which  are  noted  in  the  Data  and  Methods  section.  A  logistic  

regression  is  used  to  determine  the  direction  and  significance  of  these  relationships.    

 

DATA  and  METHODS  

 

  The  data  used  in  the  research  are  taken  from  the  BRFSS  (  Behavioral  Risk  

Factor  Surveillance  Survey)  for  the  year  2007.    The  BRFSS  is  a  random-­‐digit-­‐dialed  

telephone  survey  of  the  non-­‐institutionalized  U.S  population  aged  18  years  and  

older.  The  telephone  numbers  represent  a  sample  of  the  population  of  households  

with  a  telephone.  It  is  a  state-­‐based  health  survey  administered  by  the  Center  for  

Disease  Control  and  covers  all  50  states,  the  District  of  Colombia,  Puerto  Rico,  Guam  

and  the  Virgin  Islands.  Although  the  data  are  available  with  post  stratification  

weights,  this  research  will  use  an  unweighted  sample.  Several  questions  from  the  

survey  were  used  to  construct  variables  for  physical  activity,  smoking  habits,  

drinking  habits,  and  visits  to  a  physician.  The  data  set  has  430,912  participants,  

Page 7: Physical Activity and Health Risks

  7  

after  excluding  respondents  with  incomplete  or  missing  data  on  physical  activity  

and  other  variables  in  the  study,  the  number  of  observations  in  the  analytical  

sample  was  reduced  to  268,  522.  The  large  number  of  missing  observations  is  

referenced  in  the  discussion  section  of  this  paper.    

 

Physical  Activity  

    Four  categories  were  constructed:  none,  low,  moderate,  and  high.  The  first  

question  asked  in  the  physical  activity  section  is    “  In  a  usual  week  ,  do  you  do  

moderate  activities  for  at  least  10  minutes  at  a  time,  such  as  brisk  walking,  bicycling,  

vacuuming,  gardening,  or  anything  else  that  causes  some  increase  in  breathing  or  

heart  rate?”    Those  who  answered  “yes”  were  then  asked  how  many  minutes  they  

are  moderately  active  for  a  day,  and  how  many  days  per  week  are  they  moderately  

active.  If  a  respondent  answered  “no”  they  were  then  directed  to  the  question  “  In  a  

usual  week  do  you  do  vigorous  activities  for  at  least  10  minutes  at  a  time  such  as  

running,  aerobics,  heavy  yard  work  or  anything  else  that  causes  large  increases  in  

breathing  or  heart  rate?”  Those  who  answered  “no”  to  both  of  these  questions  are  

defined  as  “None”.    Those  who  answered  they  engage  in  vigorous  activity  were  

asked  how  many  days  per  week  do  they  engage  in  this  activity  as  well  as  how  much  

time  they  spend  doing  these  activities  on  those  days.  The  number  of  minutes  and  

days  were  then  multiplied.    The  CDC  recommends  that  a  person  engage  in  30  

minutes  of  activity  preferably  every  day  of  the  week,  but  at  least  on  most  days.  

Those  who  did    a  total  of  moderate  activity  for  120  –  210  minutes  per  week  are  

categorized  as  “Moderate”  meaning  they  meet  the  recommended  level.  Those  who  

answered  they  engage  in  moderate  activity  for  less  than  120  minutes  are  

categorized  as  “Low”  and  those  who  engage  in  moderate  activity  for  over  210  

minutes  per  week  are  categorized  as  “High”.  Those  who  answered  that  they  engage  

in  vigorous  activity  for  50  –  100  minutes  are  categorized  as  “  Moderate”,    those  who  

engage  in  vigorous  activity  for  less  than  50  minutes  per  week  are  “  Low”  and  those  

engaging  in  over  100  minutes  are  “High”.  The  construction  of  these  variables  is  

based  on  other  research  which  used  similar  definitions  and  measures  (Brown  

2003,Brown  2004,Sherwood  2000).    

Page 8: Physical Activity and Health Risks

  8  

 

Annual  Checkup  

    To  determine  how  often  a  respondent  visits  a  doctor  the  question  

asked  was  “  About  how  long  has  it  been  since  you  last  visited  a  doctor  for  a  routine  

checkup?  A  routine  checkup  is  a  general  physical  exam    not  an  exam  for  a  specific  

injury  ,  illness,  or  condition.”  With  the  options  to  answer  “  Within  the  past  year,  

Within  the  past  2  years,  Within  the  past  5  years,  5  or  more  years  ago.”    For  the  

purpose  of  this  research  respondents  are  grouped  as  either  visiting  the  doctor  

within  the  past  year,  or  if  it  has  been  longer  than  one  year  since  their  last  checkup.    

 

Smoking  

  Respondents  were  put  into  two  categories  for  smoking  :  Smoker  or  Non-­‐

Smoker.  Non-­‐Smokers  are  those  who  have  smoked  less  than  100  cigarettes  in  their  

entire  life  or  those  who  have  smoked  more  than  100  cigarettes  but  currently  do  not  

smoke.  Those  who  answered  they  currently  smoke  are  classified  as  smokers.  Those  

who  said  they  have  ever  smoked  more  than  100  cigarettes,  but  are  not  current  

smokers  were  categorized  as  non-­‐smokers.  Other  literature  has  shown  that  ex-­‐

smokers  have  better  quality  of  life  and  significant  improvement  in  respiratory  

symptoms,  effects  from  long  term  disease,  as  well  as  other  health  benefits  that  are  

generally  associated  with  non-­‐smokers  (Rothenberg  1990).  

Drinking  

  Respondents  were  first  asked  “During  the  past  30  days  have  you  had  at  least  

one  drink  of  any  alcoholic  beverage  such  as  beer,  wine,  a  malt  beverage  or  liquor?”  

Those  who  answered  “No”  were  put  into  the  “Non-­‐Drinker”  category.  Those  who  

report  at  least  one  alcoholic  drink  within  the  past  30  days  were  then  asked  “During  

the  past  30  days  how  many  days  per  week  did  you  have  at  least  one  drink  of  any  

alcoholic  beverage?”  as  well  as  “One  drink  is  equivalent  to  a  12-­‐ounce  beer,  a  5-­‐

ounce  glass  of  wine,  or  a  drink  with  one  shot  of  liquor.  During  the  past  30-­‐days,  on  

days  when  you  drank,  about  how  many  drinks  did  you  drink  on  average?”  To  

determine  the  number  of  drinks  a  respondent  had  in  one  week  the  number  of  days  

they  said  they  drink  was  multiplied  by  the  average  number  of  drinks  they  had  in  a  

Page 9: Physical Activity and Health Risks

  9  

sitting.  For  females  moderate  alcohol  consumption  is  considered  to  be  one  drink  per  

day,  and  2  drinks  per  day  for  males.  Women  who  drank  1  to  7  drinks  per  week  are  

categorized  as  “  Moderate”  and  Men  who  drank  1  to  14  drinks  per  week  are  

categorized  as  “Moderate.”  Females  drinking  more  than  7  and  Males  drinking  more  

than  14  are  categorized  as  “Heavy.”  

 

 

Analysis  

   A  logistic  regression  was  used  to  examine  the  relationship  between  annual  

checkups  and  smoking.  Ordered  logit  was  used  to  examine  drinking  and  activity  

levels.  A  chi-­‐square  was  run  on  activity  and  demographic  variables  to  examine  

associations.    Table  1  displays  descriptive  statistics  from  the  Chi-­‐Square.  The  

number  of  observations  for  each  activity  level  by  demographic  characteristics  is  

shown.  Table  2  displays  the  odds  ratios  for  activity  level  and  doctor  visit  within  the  

past  year.  The  first  model  does  not  control  for  any  variables,  Model  2  through  Model  

4  controls  for  self-­‐reported  health,  age,  sex,  race,  income  and  education.  Table  3  

displays  odds  ratios  for  activity  level  and  smoking.  Once  again  the  first  model  has  no  

controls.  Model  2  through  Model  4  implement  control  variables.  Table  4  displays  

odds  ratios  for  activity  level  and  drinking.  The  models  follow  the  same  pattern  as  

the  previous  tables.      

  As  demonstrated  in  the  tables,  four  models  were  run.  The  first  model  does  

not  control  for  any  other  variables  and  the  results  shown  are  just  for  activity  levels  

vs.  the  dependent  variable  (annual  checkup,  smoking,  and  drinking)  .  Model  2  

controls  for  a  person’s  self  evaluated  health  and  the  reference  group  is  ‘Excellent’.  

Model  3  controls  for  self  evaluated  health,  age,  sex,  and  race.  The  reference  groups  

are  18-­‐34  for  age  (  the  youngest  age  group),  white  for  race  and  males  for  sex.  Model  

4  controls  for  all  of  the  previous  including  Income  and  Education.  The  reference  

groups  are  less  than  $10,000  for  income  and  Less  than  High  School  for  education.    

 

 

 

Page 10: Physical Activity and Health Risks

  10  

 

RESULTS  

  Table  1  through  Table  3  show  percentage  distributions  of  the  dependent  

variables.  Table  1  displays  medical  exams  by  smoking  status.  The  percentages  of  

smokers  who  have  seen  a  doctor  within  the  past  year  is  lower  than  the  percentage  

of  non-­‐smokers  who  have  a  seen  a  doctor  within  the  past  year.  Table  2  displays  

medical  exams  by  drinking  status.  Non-­‐  drinkers  have  the  highest  percentage  of  

those  who  have  seen  a  doctor  within  the  past  year.  The  percentage  drops  for  

moderate  and  more  for  heavy  drinkers.  These  two  tables  show  that  those  who  do  

not  drink  or  smoke  display  the  highest  percentages  for  receiving  a  medical  exam  

within  the  past  year.    

  Table  3  displays  smoking  status  by  drinking  status.  This  table  shows  that  

among  drinkers,  the  highest  percentage  of  smokers  is  found  in  the  ‘heavy  drinking’  

category.  The  percentage  of  smokers  in  the  ‘heavy  drinking’  category  is  double  the  

percentage  of  smokers  in  the  ‘none’  and  ‘moderate’  drinking  categories.  This  

suggests  that  smoking  and  heavy  drinking  may  go  hand  in  hand  with  each  other.    

Table  4  displays  a  cross  classification  of  activity  level  by  all  of  the  dependent  

variables  in  the  analysis.  The  Chi-­‐Square  tests  showed  a  significant  in  relationship  

with  each  variable  with  a  p-­‐value  of  <.001.  Eleven  percent  of  Males  are  meeting  the  

‘moderate’  level  of  activity  while  28%  are  meeting  the  requirements  for  ‘high’  

activity  levels.  Nine  percent  of  Females  are  meeting  the  ‘moderate’  level  while  17%  

are  meeting  the  ‘high’  level.    This  shows  us  that  Males  are  more  active  than  Females.  

Since  the  data  are  unweighted  we  see  a  sample  that  is  predominantly  female.  Thirty  

five  percent  of  the  sample  is  male  and  65%  is  female.  Whites  and  ‘Other’  races  are  

the  most  active,  followed  by  Hispanics  and  Blacks.  Younger  age  groups  are  the  most  

active,  activity  declines  as  age  increases.  Those  with  higher  income  and  education  

have  higher  activity  levels  than  those  with  lower  income  and  education.    

 

Annual  Checkup  (  Table  5)  

  Activity  is  significantly  related  to  having  had  a  medical  exam  within  the  past  

year  (  p-­‐value  <  .0001)  when  no  other  variables  are  controlled.  Compared  to  an  

Page 11: Physical Activity and Health Risks

  11  

activity  level  of  ‘None’,  all  other  activity  levels  are  less  likely  to  have  seen  a  doctor  

within  the  past  year.  As  level  of  activity  increases,  it  is  increasingly  less  likely  that  

one  has  seen  a  doctor  within  the  past  year.  When  controlling  for  self  reported  

health,  this  pattern  remains.  Model  3  begins  to  control  for  demographic  variables.  In  

Model  3  activity  level  ‘Low’  is  no  longer  significant  when  compared  to  

‘None’.’Moderate’  and  ‘High’  remain  significant,  however  they  change  directions.  

‘Moderate’  and  ‘High’  now  become  more  likely  than  those  with  no  activity  to  have  

seen  a  doctor  within  the  past  year.  When  all  variables  are  controlled,  only  one  

activity  level  remains  significant  when  compared  to  ‘None’.    ‘High’  activity  levels  are  

significant  at  level  .03.  Those  with  ‘High’  activity  level  compared  to  those  of  ‘None’  

are  less  likely  to  have  seen  a  doctor  within  the  past  year.  Activity  levels  ‘Low’  and  

‘Moderate’  have  no  significant  difference  when  compared  to  ‘None’.    

  Those  who  reported  ‘Poor’  health  are  the  most  likely  to  go  to  the  doctor  

within  a  year  as  compared  to  those  with  ‘Excellent’  health’.    As  self-­‐reported  health  

goes  from  ‘Poor’  to  ‘Excellent’,  the  chances  of  seeing  a  doctor  within  a  year  become  

less.  Females  are  more  likely  to  have  seen  a  doctor  within  the  past  year  than  males,  

older  age  groups  are  more  likely  to  have  a  seen  a  doctor  within  the  past  year  than  

younger  age  groups.  All  race  groups  are  more  likely  than  Whites  to  have  seen  a  

doctor  within  the  past  year  with  Blacks  being  the  most  likely,  followed  by  Hispanics  

and  then  “Other”.  As  education  rises  so  does  the  likelihood  of  seeing  a  doctor  within  

the  past  year.  Incomes  of  $25,000  or  more  are  significant,  as  income  rises  so  do  the  

chances  of  one  seeing  a  doctor  within  the  past  year.  

  Activity  level  and  doctor  visits  are  significant.  The  directions  between  the  

reference  group  and  levels  of  activity  change  based  on  which  demographics  are  

being  controlled.  The  result  of  controlling  for  all  demographic  variables  is  that  only  

those  with  a  sedentary  lifestyle  compared  to  those  with  the  highest  level  of  activity  

show  significance  differences.  Those  who  are  sedentary  are  more  likely  to  have  seen  

the  doctor  within  the  past  year  as  opposed  to  those  who  exhibit  the  highest  physical  

activity  level.    

 

Page 12: Physical Activity and Health Risks

  12  

 

 Smoking (Table 6)

Model 1 shows that when no variables are controlled, activity is significantly

related to being a non-smoker (p-value <.001). Those with an activity level of ‘Low’ or

‘High’ are most likely to be non-smokers; ‘Moderate’ activity is least likely to smoke

compared to ‘None’. When controlling for self-reported health, activity level ‘Low’ is no

longer significant compared to ‘None’. ‘Moderate’ level activity is still significant with a

less likely chance of smoking, and ‘High’ remains significant but changes its effect,

and those within the group become more likely to smoke than the ‘None’ group. Model

3 controls for age, sex, and race. Activity becomes significant at all levels again and

‘Low’, ‘Moderate’ and ‘High’ are all less likely to smoke than ‘None’. When all

variables are held constant the only activity level significant to ‘None’ is

‘Moderate’. ‘Moderate’ is less likely than ‘None’ to be a smoker. ‘Low’ and ‘High’

show non-significant p-values (.1708,.7422) when compared to ‘None’. As self-reported

health becomes worse, the odds of smoking become greater. Females are less likely to

smoke than males, blacks and Hispanics are less likely to smoke than whites, and all age

groups are less likely to smoke than 18-34 year olds. The 35-49 year olds, however, are

not significantly different from 18-34 year olds as being a smoker vs. non-smoker in

Model 3. Moving to Model 4, the p-value becomes significant (<.0001) suggesting that

35-49 year olds are more likely to smoke than 18-34 year olds. Education and income

are both significant with all levels of education being less likely to smoke than ‘Less than

High School’ as well as all levels of income being less likely to smoke than ‘ Less than

$10,000’. As both income and education rise the odds of smoking become less.

Physical activity and smoking are significant when no demographic variables are

controlled . Once age, sex, race, income, and education are controlled ‘Moderate’, it is

the only activity level which remains significant to the reference group of ‘None’. Having

‘Low’ activity or ‘High’ activity is not significant to a sedentary lifestyle when

examining whether a person is a smoker or non-smoker.

Page 13: Physical Activity and Health Risks

  13  

 

 

Drinking  (  Table  7)  

  When  no  variables  are  controlled,  all  levels  of  activity  are  significant  with  p-­‐

value  <.001.  An  ordered  logit  shows  that  as  activity  levels  rise  so  do  the  odds  of  

being  a  moderate  or  heavy  drinker  as  opposed  to  not  drinking.  When  controlling  for  

self-­‐reported  health  this  pattern  remains.  Activity  level  and  drinking  remain  

significant  across  each  model.  All  activity  groups  have  greater  odds  of  being  

moderate  or  heavy  drinkers  compared  to  ‘None’.  As  activity  levels  rise,  so  do  the  

odds  of  being  a  drinker.    

   Self-­‐reported  health  is  also  significant  and  as  health  increases  from  ‘Poor’  to  

‘Excellent’  ,  the  odds  of  being  a  moderate  or  heavy  drinker  increase.  Sex,  race,  age,  

income  and  education  are  all  significant  as  well.  Females  are  less  likely  than  males  

to  be  moderate  or  heavy  drinkers,  all  races  are  less  likely  than  Whites  to  be  

moderate  or  heavy  drinkers.  As  both  income  and  education  rise,  the  ods  of  being  a  

moderate  or  heavy  drinker  rise.    When  income  and  education  are  not  controlled  for  

age  groups  35-­‐49  and  50-­‐64  have  higher  odds  of  being  moderate  or  heavy  drinkers  

compared  to  18-­‐34  year  olds.  65+  have  lower  odds  than  18-­‐34  year  olds  to  be  

moderate  or  heavy  drinkers.    In  Model  4  when  these  variables  are  controlled  for  all  

age  groups  have  less  odds  of  being  moderate  of  heavy  drinkers  compared  to  18-­‐34  

year  olds,  however  there  is  not  a  significant  difference  between  18-­‐34  year  olds  and  

50-­‐64  year  olds.    

  Activity  level  and  drinking  are  significant  at  all  levels  across  all  models.  

Higher  activity  levels  display  higher  odds  of  being  a  moderate  or  heavy  drinker  

compared  to  a  non-­‐drinker.  This  pattern  remains  constant  across  all  controls  of  age,  

sex,  race,  income,  education  and  self-­‐reported  health.    

     

DISCUSSION     This  research  has  demonstrated  that  a  person’s  level  of  activity  can  impact  

other  health  behaviors,  although  it  may  not  be  in  the  direction  originally  

hypothesized.  There  is  a  significant  relationship  between  activity  level  and  doctor  

Page 14: Physical Activity and Health Risks

  14  

visits.  When  controls  are  set,  the  significance  is  only  between  the  two  extreme  

groups  of  activity  levels,  ‘None’  compared  to  ‘High’.  Those  who  are  of  ‘Low’  or  

‘Moderate’  do  not  show  enough  significance  to  differ  from  those  living  a  sedentary  

lifestyle.  The  conclusion  is  that  those  who  are  sedentary  have  higher  odds  of  seeing  

a  doctor  within  the  past  year  than  those  who  maintain  the  highest  level  of  physical  

activity.    

  Smoking  and  activity  level  show  significance  with  all  levels  when  no  control  

variables  are  set  in  place.  Once  demographics  are  controlled  for  only  one  group  is  

significant  to  the  reference  group.  ‘Moderate’  activity  level  is  the  only  significance  

compared  to  ‘None’.  Those  who  have  ‘Low’  and  ‘High’  activity  levels  show  no  

significance  to  being  a  non-­‐smoker  versus  a  smoker  when  being  compared  to  ‘None’.  

Only  those  with  ‘Moderate’  activity  level  are  less  likely  to  smoke.    

  All  activity  levels  are  significant  to  drinking.  Every  level  remains  significant  

through  each  model  as  demographics  are  controlled  for.  As  the  level  of  activity  

increases,  so  do  the  odds  of  being  a  moderate  or  heavy  drinker  as  opposed  to  not  

drinking.  Although  this  does  not  meet  the  original  hypothesis,  there  is  still  

significance  between  activity  levels  and  drinking  it  is  just  shown  in  a  different  

direction  than  what  was  expected.    

  Finding  this  direction  between  drinking  and  physical  activity  brings  up  some  

questions.  Since  we  see  those  who  are  in  the  highest  activity  level  also  being  the  

most  likely  drink,  is  drinking  really  considered  as  risky  a  health  behavior  as  

originally  thought?  Is  there  the  possibility  that  moderate  consumption  may  be  

healthy  which  may  be  why  we  see  those  in  the  highest  activity  level  having  the  

highest  odds  of  moderate  or  heavy  drinking  compared  to  non-­‐drinking?  Some  

previous  studies  have  suggested  a  protective  effect  from  drinking  light  to  moderate  

amounts  of  alcohol.  They  have  indicated  that  that  moderate  alcohol  consumption  

can  protect  against  cardiovascular  disease.  The  effect  has  been  demonstrated  across  

all  age  ranges  and  is  present  for  both  men  and  women  (  Thakker,  1998).  

   Future  research  should  examine  age  groups  across  a  lifetime.  This  research  

examined  age  groups  during  the  same  time  period.  Examining  how  a  certain  age  

cohort’s  activity  and  health  habits  change  as  they  move  from  18-­‐34  to  65+  could  

Page 15: Physical Activity and Health Risks

  15  

prove  to  be  significant  for  future  researchers.  It  is  also  important  to  note  the  BRFSS  

has  several  limitations  which  are  common  among  most  surveys.  Certain  problems  

may  arise,  due  to  inaccuracy  or  respondents  recalling  their  behavior.  

  One  limitation  to  reference  about  this  research  in  particular  is  the  large  

number  of  missing  data.  The  data  set  originally  had  430,912  observations  that  were  

cut  down  to  268,522.  This  was  caused  by  the  number  of  participants  who  did  not  

answer  certain  questions  that  were  examined.  The  largest  number  came  from  

58,622  who  did  not  report  their  household  income.  Those  who  did  not  report  

household  income  were  not  included  in  the  research.  The  observations  in  which  

income  was  not  reported  had  higher  percentages  of  being  in  the  ‘None’  activity  level  

category  than  those  who  did  report  income.  Table  8  displays  the  descriptive  

statistics  for  those  cases  which  did  not  report  their  income.  These  percentages  show  

that  for  activity,  the  largest  percentage  missing  were  ‘None’.  Those  who  had  seen  a  

doctor  within  the  past  year,  smokers,  and  heavy  drinkers,  were  all  the  most  likely  to  

be  excluded  from  this  analysis  because  they  did  not  report  their  household  income.  

Age  group  65  plus,  females,  Hispanics,  and  those  with  less  than  a  high  school  

education  were  all  most  likely  to  be  excluded  due  to  missing  data  for  household  

income.    

  The  research  suggests  that  physical  activity  can  predict  health  behaviors.  The  

findings  show  those  who  are  less  active  are  more  likely  to  have  received  a  medical  

exam  within  the  past  12  months,  the  moderately  active  are  least  likely  to  smoke,  and  

those  who  are  the  most  active  are  also  the  most  likely  to  be  a  moderate  or  heavy  

drinker.    Although  this  research  alone  cannot  say  higher  level  of  physical  activity  

produce  better  health  habits,  combined  with  the  previous  literature  it  does  point  us  

in  the  direction  of  seeing  that  being  physically  active  can  have  positive  effects  on  

health  habits  and  healthy  lifestyles.    

 REFERENCES    

1) Blair  SN,  Kampert  JB,  Kohl  HW,  Barlow  CE,  Macere  CA,  et  al.  Influences  of  cardiorespiratory  fitness  and  other  precursors  on  cardiovascular  disease  and  all-­‐cause  mortality  in  men  and  women.  JAMA.1996;276  (3)  :  205-­‐10.  

Page 16: Physical Activity and Health Risks

  16  

2) Brawley,  Lawrence  R.,  Rejeski,  W.Jack,  King,  Abbey  C.  Promoting  Physical  Activity  for  Older  Adults;  the  challenges  for  changing  behavior.  Am  J  Prev  Med  2003;25(172-­‐183).    

3) Brett,  Kate  M.  &  Burt,  Catherine  W.  Utilization  of  Ambulatory  Medical  Care  by  Women:  United  States  1997-­‐98.  2001.  Vital  and  Health  Statistics  13(149).  

4) Brown,  David  W.,  Balluz,  Lina  S.,  Heath,  Gregory  W.,  Moriarty,  David  G.,  Ford,  Earl  S.,  Giles,  Wayne  H.,  Mokdad,  Ali  H.  Associations  between  recommended  levels  of  physical  activity  and  health-­‐related  quality  of  life  Finding  from  the  2001  Behavioral  Risk  Factor  Surveillance  System  BRFSS  survey.  Science  Direct.  2003  530-­‐528.  

5) Brown,  David  W.,  Brown,  David  R.,  Heath,  Gregory  W.,  Balluz,  Lina  S.,  Giles,  Wayne  H.,  Ford,  Earl  S,  Mokdad,  Ali  H.  Associations  between  physical  activity  dose  and  health  related  quality  of  life.  Medicine  and  Science  in  Sports  and  Exercise.  2004.  890-­‐896.  

6) Caspersen  CJ,  Merrit  RK,  Health  GW,  Yeager  KK.  Physical  activity  patterns  of  adults  aged  60  years  and  older.  Med.  Sci.    Sports  Exc.  22  (suppl):S79.    

7) Caspersen  CJ,  Merritt  RK.  Physical  activity  trends  among  26  states,  1986-­‐1990.  1995.  Med.  Sci.  Sports  Exerc,  27(5):  713-­‐20.    

8) Caspersen  CJ,  Merritt  RK.  Trends  in  physical  activity  patterns  among  older  adults:  The  Behaviroal  Risk  Factors  Surveillance  System,  1986-­‐1990.  1992;  Med.  Sci  Sports  Exerc.  24(4):526  

9) Centers  for  Disease  Control  and  Prevention.  Annual  smoking-­‐attributable  mortality,  years  of  potential  life  lost  and  economic  costs.  United  States,  1995-­‐1999.  MMWR  Morb  Mortal  Wkly  Rep.  2002;  51:  300-­‐303.  

10) Centers  for  Disease  Control  and  Prevention.  Neighborhood  safety  and  the  prevalence  of  physical  activity  –  selected  states.  ,  1996  Morb  Mortal  Wkly  Rep  1999;48:143-­‐6.    

11) Centers  for  Disease  Control  and  Prevention.  Cigarette  Smoking-­‐attributable  morbidity  –  United  States,  2000.  MMWR  Mob  Mortal  Wkly  Rep.  2003;  52:  842-­‐844.  

12) Danaei,  Goodarz,  Ding,  Eric  L.,  Mozaffarian,  Dariush,  Taylor,  Ben,  Rehm,  Jurgen,  Murray,  J.L.  Christopher,  Ezzati,  Majid.  The  preventable  causes  of  death  in  the  united  states:  comparative  risk  assessment  of  dietary,  lifestyle,  and  metabolic  risk  factors.  PLoS  Med  2009.  6(4)1-­‐22.  

13) Fiore  MC,  Novotny  TE,  Pierce  JP  et  al.  Trends  in  cigarette  smoking  in  the  United  States.  The  changing  influence  of  gender  and  race.  JAMA  1989;  261:  49-­‐55.  

14) Flint  AJ,  Novonty  TE.  Trends  in  black/white  differences  in  current  smoking  among  18-­‐25  year  olds  in  the  United  States.  1983-­‐1993.  Am  J  Prev  Med.  1998;  14:  19-­‐24.  

15) Hughes,  J.R.  Psychological  Effects  of  Habitual  Aerobic  Exercise:  A  Critical  Review.  Prev  Med  13:  66-­‐78  (1984).  

16) Keife  CI,  Williams  OD,  Lewis  CE  et  al.  Ten-­‐year  changes  in  smoking  among  young  adults:  are  racial  differences  explained  by  sociodemographic  factors  in  CARDIA  study?  Am  J  Public  Health.  2001;  91:213-­‐218  

Page 17: Physical Activity and Health Risks

  17  

17) McAuley  E.  Physical  activity  and  psychosocial  outcomes.  In  Bouchard  C,  Sherperd  RJ,  Stephens  T,  eds.  Physical  activity,  fitness,  and  health.  Champaign  IL,  Health  and  Kinetics  1994;  551-­‐68  

18) McGinnis  JM  Foege,  WH.  Actual  causes  of  death  in  the  United  States.  JAMA.  1993;  270:  2207-­‐2212  

19) Merrick,  Elizabeth  L.  ,  Horgan,  Constance  M.  ,  Hodgkin,  Dominic,  Garnick,  Deborah  W.,  Houghton,  Susan  F.,  Panas,  Lee  ,  Saitz,  Richard  ,  Blow,  Fredric  C.  Unhealthy  drinking  patterns  in  older  adults  :  prevalence  and  associated  characteristics.  JAGS.  2008;  56:214-­‐223.    

20) Nam,  Charles  B.,  Robert  A.  Hummer,  and  Richard  G.  Rogers.  Underlying  and  Multiple  Causes  of  Death  Related  to  Smoking.  1994.  Population  Research  and  Policy  Review.  13:  305-­‐325.  

21) National  Center  for  Health  Statistics,  Health  Data  Interactive,  www.cdc.gov/nchs/hdi.htm.  Accessed  on  July  19,  2011.    

22) Penedo,  Frank  J.  &  Dahn,  Jason  R.  Exericse  and  well-­‐being  :  a  review  of  mental  and  physical  health  benefits  associated  with  physical  activity.  2005.  Current  Opinion  in  Psychiatry.  18:189-­‐193.      

23) Pierce  JP,  Fiore  MC,  Novotny  TE  et  al.  Trends  in  cigarette  smoking  in  the  United  States.  Educational  differences  are  increasing.  JAMA.  1989;  261:  56-­‐60.    

24) Rogers  Richard  G.  and  Eve  Powell-­‐Grinner.  Life  expectancies  of  cigarette  smokers  in  the  United  States.  1991.  Social  Science  and  Medicine.32(10):1151-­‐1159.    

25) Rothenberg  RG,  Koplan  JP.  Chronic  disease  in  the  1990s.  Ann  Rev  Public  Health  1990;11  267-­‐96.      

26) Sherwood,  Nancy  &  Jeffery,  Robert  W.  The  Behavioral  Determinants  of  Exercise:  Implications  for  physical  activity  interventions.  Annual  Reviews.  2000.  20:21-­‐44.    

27) Slesinger,  Doris  P.,  Tessler,  Richard  C.,  Mechanic  ,  David.  The  effects  of  social  characteristics  on  the  utilization  of  preventive  medical  services  in  contrasting  health  care  programs.  Medical  Care.  1976.  14:  5;392-­‐404.    

28) Tillman,  Maria  &  Silcock  Jonathan.  A  comparison  of  smokers;  and  ex-­‐smokers’  health-­‐related  quality  of  life.  Journal  of  Public  Health  Medicine.  1997;19(3):268-­‐273.    

29) Thakker,  Kerstin  Damstron.  An  overview  of  health  risks  and  benefits  of  alcohol  consumption.  Alcoholism:  Clinical  and  Experimental  Research.  1998.  22(7)  285-­‐298.  

 30) Trost,  S.G,  Owen,  N.,  Bauman,  Adrian  E.,  Sallis,  James  F.,  Brown,  Wendy.  Correlated  of  

adult’s  participation  in  physical  activity:  review  and  update.  Med  Sci  Sports  Exerc.,  2002.  Vol  34,  No.  12,  pp  1996  –  2001.  

Page 18: Physical Activity and Health Risks

  18  

31) U.S.  Department  of  Health  and  Human  Services  Physical  Activity  and  Health:  A  Report  of  the  Surgeon  General  Atlanta  (GA):  U.S.  Department  of  Health  and  Human  Services,  Centers  for  Disease  Control  and  Prevention,  National  Center  for  Chronic  Disease  Prevention  and  Health  Promotion.;  1996.  

32) Warburton,  Darren  E.R.,  Nicol,  Crystal  Whitney,  Bredin,  Shannon  S.D.  Health  benefits  of  physical  activity:  the  evidence.  CMAJ.  2006;  Vol.  174  No.6.    

33) Weschler  H,  Austin  SB.  Binge  drinking  the  five/four  measure.  J  Stuf  Alcohol.  1998;  59:  122  -­‐124.  

34) Weschsler    H,  Nelson  TF.  Binge  drinking  and  the  American  college  student:  what’s  five  drinks?  Psychol  Addict  Behav.  2001;  15:  287-­‐291.      

                                                             

Page 19: Physical Activity and Health Risks

  19  

 TABLE  1.  Percentage  Distribution  of  Medical  Exam  by  Smoking    

   

Non-­‐Smoker   Smoker  

Within  Year  

 76.27   65.94  

Longer  than  1  Year   23.73   34.06  

     BRFSS,  2007  

     

TABLE  2.  Percentage  Distribution  of    Medical  Exam  by  Drinking        

   

Non-­‐Drinker  

Moderate  Drinker  

Heavy  Drinker  

Within  Year  

 75.65   71.53   66.19  

Longer  than  1  Year   24.35   28.48   33.81    

BRFSS,  2007        

TABLE  3.  Percentage  Distribution  of    Smoking    by  Drinking    

   

Non-­‐Drinker  

Moderate  Drinker  

Heavy  Drinker  

Non-­‐Smoker  

 83.68   83.66   65.36  

Smoker    

16.32   16.34   34.64    

BRFSS,  2007  

                   

Page 20: Physical Activity and Health Risks

  20  

 Table  4.  Percent  Distribution  of  Physical  Activity  by  Social  Demographics  

 

            ACTIVITY              

     None   Low   Moderate   High  

Sex       N                  

 Male   92,633   15.01   45.42   10.97   28.6  

    Female   175,889   17.86   55.78   8.95   17.42  Race  

                White   206,728   15.42   52.59   10.05   21.93  

 Black   23,589   23.1   52.27   7.93   16.7  

    Hispanic   21,139   24.09   49.77   7.81   18.34  

 Other   17,066   17.02   50.43   9.33   23.22  

Age                          

 18-­‐34   34,583   9.9   45.27   14.15   30.68  

    35-­‐49   69,988   11.78   48.53   12.89   26.8  

 50-­‐64   77,600   16.03   53.66   9.6   20.7  

    65+   84,106   24.78   56.56   5.19   13.47  Income  

                less  than  $10,000   14,575   33.13   51.91   4.36   10.6  

 $10,000  -­‐  $24,999   60,504   24.12   54.84   6.3   14.74  

    $25,000  -­‐  $49,999   30,745   16.94   54.42   8.64   20  

 $50,000  -­‐  $74,999   72,890   11.05   52.21   12.01   24.73  

    $75,000+   51,208   6.88   45.89   14.61   32.63  Education  

             

less  than  High  School   32,797   31.36   51.47   4.76   12.4  

 High  School   86,119   19.79   54.15   7.8   18.26  

    Some  College   68,314   14.83   53.25   10.08   21.84       College  Graduate   80,353   9.52   49.34   13.33   27.81    

 BRFSS,2007  

                 

Page 21: Physical Activity and Health Risks

  21  

 TABLE  5.  Odds  Ratios  from  Logistic  Regression  on  Activity  Level  and    

Medical  Visits  in  Last  Year  Medical  Visits  in  Last  Year    

 

     BRFSS,2007  

           

 

   Model  1   p-­‐value   Model  2   p-­‐value   Model  3   p-­‐value   Model  4   p-­‐value  

Activity   Low   0.819   <.0001   0.918   <.0001   1.013   0.3526   0.978   0.1525  

 Moderate   0.652   <.0001   0.786   <.0001   1.06   0.0024   0.986   0.5023  

 High   0.638   <.0001   0.773   <.0001   1.038   0.0228   0.963   0.0349  

General  Health   Poor  

   1.961   <.0001   1.543   <.0001   1.922   <.0001  

 Fair  

   1.617   <.0001   1.277   <.0001   1.517   <.0001  

 Good  

   1.265   <.0001   1.10   <.0001   1.211   <.0001  

 

Very  Good  

   1.116   <.0001   1.06   <.0001   1.086   <.0001  

Age   18-­‐34          

1.00    

1.00    

 35-­‐49  

       1.16   <.0001   1.101   <.0001  

 50-­‐64  

       1.862   <.0001   1.1814   <.0001  

 65+  

       3.95   <.0001   3.967   <.0001  

Sex   Female          

1.433   <.0001   1.50   <.0001  Race   Black  

       1.888   <.0001   2.111   <.0001  

 Hispanic  

       1.138   <.0001   1.3   <.0001  

 Other  

       1.045   0.0185   1.092   <.0001  

Education   High  School            

1.058   0.0021  

 Some  College  

         1.039   0.0492  

 College  Gradute  

         1.069   0.0008  

Income   $10,000  -­‐  $24,999            

0.983   0.4633  

 $25,000  -­‐  $49,999  

         1.124   <.0001  

 $50,000  -­‐  $74,999  

         1.412   <.0001  

 $75,000  +  

           1.775   <.0001  

Page 22: Physical Activity and Health Risks

  22  

       

TABLE  6.  Odds  Ratios  from  Logistic  Regression  on  Physical  Activity  and  Smoking    

   Model  1   p-­‐value   Model  2   p-­‐value   Model  3   p-­‐value   Model  4   p-­‐value  

Activity   Low   0.857   <.0001   1.008   0.5921   0.895   <.0001   0.978   0.1708  

 Moderate   0.688   <.0001   0.915   <.0001   0.678   <.0001   0.822   <.0001  

 High   0.827   <.0001   1.121   <.0001   0.854   <.0001   1.007   0.7422  

General  Health   Poor  

   3.006   <.0001   3.995   <.0001   2.278   <.0001  

 Fair  

   2.267   <.0001   3.041   <.0001   1.927   <.0001  

 Good  

   1.834   <.0001   2.171   <.0001   1.676   <.0001  

 

Very  Good  

   1.442   <.0001   1.528   <.0001   1.403   <.0001  

Age   35-­‐49          

0.989   0.5013   1.119   <.0001  

 50-­‐64  

       0.74   <.0001   0.814   <.0001  

 65+  

       0.279   <.0001   0.253   <.0001  

Sex   Female          

0.87   <.0001   0.825   <.0001  Race   Black  

       0.504   <.0001   0.615   <.0001  

 Hispanic  

       0.99   <.0001   0.362   <.0001  

 Other  

       0.87   0.6277   0.935   0.0035  

Education   High  School            

0.784   <.0001  

 Some  College  

         0.703   <.0001  

 College  Graduate  

         0.357   <.0001  

Income   $10,000  -­‐  $24,999            

0.846   <.0001  

 $25,000  -­‐  $49,999  

         0.679   <.0001  

 $50,000  -­‐  $74,999  

         0.526   <.0001  

 $75,000  +  

           0.348   <.0001  

 BRFSS,  2007                    

   

Page 23: Physical Activity and Health Risks

  23  

 TABLE  7.  Odds  Ratios  from  Ordered  Logit  on  Activity  Level  and  Drinking    

   Model  1   p-­‐value   Model  2   p-­‐value   Model  3   p-­‐value   Model  4   p-­‐value  

Activity   Low   2.213   <.0001   1.708   <.0001   1.608   <.0001   1.417   <.0001  

 Moderate   4.084   <.0001   2.656   <.0001   2.209   <.0001   1.754   <.0001  

 High   4.386   <.0001   2.819   <.0001   2.288   <.0001   1.858   <.0001  

General  Health   Poor  

   0.188   <.0001   0.196   <.001   0.335   <.0001  

 Fair  

   0.292   <.0001   0.322   <.0001   0.499   <.0001  

 Good  

   0.524   <.0001   0.55   <.0001   0.71   <.0001  

 

Very  Good  

   0.853   <.0001   0.847   <.0001   0.922   <.0001  

Age   35-­‐49          

1.176   <.0001   .984   0.3238  

 50-­‐64  

       1.031   0.0498   .891   <.0001  

 65+  

       .711   <.0001   .795   <.0001  

Sex   Female          

0.523   <.0001   0.567   <.0001  Race   Black  

       0.462   <.0001   0.579   <.0001  

 Hispanic  

       0.531   <.0001   0.757   <.0001  

 Other  

       0.546   <.0001   0.568   <.0001  

Education   High  School            

1.317   <.0001  

 Some  College  

         1.647   <.0001  

 College  Graduate  

         2.089   <.0001  

Income   $10,000  -­‐  $24,999            

1.235   <.001  

 $25,000  -­‐  $49,999  

         1.584   <.0001  

 $50,000  -­‐  $74,999  

         2.051   <.0001  

 $75,000  +  

           3.401   <.001  

 BRFSS,2007  

       

                 

Page 24: Physical Activity and Health Risks

  24  

TABLE  8.  Percent  Distribution  of  Reported  Household  Income  Data  &  Excluded  from  the  Analysis    

 Activity  

       

None   Low   Moderate   High  Reported   79.89   85.10   90.13   89.88  Not  Reported   20.11   14.90   9.87   10.12  

 Medical  Exam  

     

 Within  Year  

Longer  than  1  Year  

   Reported   84.91   88.19      Not  Reported   15.09   11.81      

 Smoker  

       

Yes   No      Reported   85.18   88.77      Not  Reported   14.82   11.23      

 Drinker  

       

No   Moderate   Heavy    Reported   84.47   90.99   92.38    Not  Reported   15.53   9.01   7.62    

 Age  

       

18-­‐34   35-­‐49   50-­‐64   65+  Reported   87.31   91.42   88.72   78.58  Not  Reported   12.69   8.58   11.28   21.42  

 Sex  

       

Male   Female      Reported   89.58   83.7      Not  Reported   10.42   16.3      

 Race  

       

White   Black   Hispanic   Other  Reported   85.97   85.75   84.29   84.5  Not  Reported   14.03   14.25   15.71   15.5  

 Less  than  HS   High  School  

Some  College  

College  Graduate  

Reported   77.74   83.59   87.61   90.17  Not  Reported   22.26   16.41   12.39   9.83  

 BRFSS,  2007  

       

Page 25: Physical Activity and Health Risks

  25