2004 ISBNPA Conference

1
A Comparison of Slowly-Evolving Versus
Rapidly-Evolving Cities on Prevalence of Obesity
and Physical Activity
Methods

Purpose This study compared the prevalence of
obesity and leisure time physical activity (PA)
in slowly-evolving versus rapidly-evolving
cities. Methods Two metropolitan statistical
areas were identified as slowly evolving New
York City and Philadelphia. Two metropolitan
statistical areas were also identified as
rapidly-evolving Dallas and Houston. The 2002
Behavioral Risk Factor Surveillance data for
obesity and leisure time PA were compared using
logistic regression modeling across
slowly-evolving and rapidly-evolving cities.
Residents in all four cities were classified as
either lean, overweight or obese and as
participating or not participating in leisure
time PA. Results Slowly-evolving cities were
associated with a a higher proportion of lean
residents (vs. obese) than rapidly-evolving
cities (odds ratio 1.53 CI 1.26-1.86, plt0.01).
Residents were 1.5 times more likely to be lean
than obese in the slowly-evolving cities as
compared to the rapidly-evolving cities.
Conclusions Results from this study suggest
that metropolitan growth rate may be an important
environmental factor related to obesity
prevalence. However, further research is needed
to better understand the relationship of growth
rate patterns and PA in the context of
transportation and built environmental supports.
The prevalence of obesity was differentially
associated with growth rate (see Figure 2).
Slowly-evolving cities were associated with a
higher proportion of lean residents (vs. obese)
than rapidly-evolving cities (odds ratio1.53 CI
1.26-1.86, plt0.01). Residents were 1.5 times more
likely to be lean than obese in the
slowly-evolving cities as compared to the rapidly
evolving cities.

• Procedures
• New York and Philadelphia (growth rate since 1950
  8.4 and 5.0, respectively). were grouped into
  slowly-evolving cities and Dallas and Houston
  (growth rate since 1950 29.3 and 25.4,
  respectively). were group into rapidly-evolving
  cities.
• The national Behavioral Risk Factor Surveillance
  System (BRFSS) was used to document body mass
  index (BMI) and leisure-time PA for the present
  study (some vs. no PA). BMI and leisure time PA
  were obtained for residents living in the
  slowly-evolving cities of New York City and
  Philadelphia (N5,143) or rapidly-evolving cites
  of Dallas and Houston (N1,864).
• BMI was calculated by self-reported data for
  weight and height (kg/m2). Residents were
  classified as lean lt 25 BMI, overweight
  25.0-29.9 BMI, or as obese gt 30 BMI.
• Analysis
• SAS-callable SUDAAN v. 8.0 was used to perform
  all analyses. Prevalence estimates of BMI and
  leisure-time PA was calculated for residents in
  each of the metropolitan statistical areas. A chi
  square statistic was used to determine the
  homogeneity of the prevalences of BMI and
  leisure-time PA between residents of
  slowly-evolving versus rapidly-evolving cites.
• Generalized logistic regressions (adjusting for
  age race and sex) were used to account for the 2
  levels of the dependent variable (type of
  evolving city) and separate models were run for
  BMI and leisure-time PA.

No significant differences were demonstrated for
leisure-time PA by metropolitan statistical areas
(see Figure 3).
Results
Introduction
Two metropolitan statistical areas were
identified as slowly evolving (see Figure 1) New
York City and Philadelphia (growth rate since
1950 8.4 and 5.0, respectively). Two
metropolitan statistical areas were also
identified as rapidly-evolving Dallas and
Houston (growth rate since 1950 29.3 and
25.4, respectively).

• But I believe, as many before me, that this is
  just the storm before the calm. The new sciences
  of chaos and complexity tell us that a system
  that is far from stable is a system ripe for
  change.
• J.M. Benyus Biomimicry
• Obesity and physical inactivity have been major
  risk factors associated with morbidity and
  mortality in the United States.
• Recently, obesity and physical inactivity have
  been on the rise. Determining connections between
  this trend and the environment could lead to a
  built environment that is conducive to healthy,
  active people.
• Rapidly evolving cities (with rapid growth rates)
  are unstable, which means that they are good
  candidates for change. We hypothesize that rising
  obesity and physical inactivity may be emerging
  properties of these cities, while cities that are
  evolving slower will not show similar changes.

Conclusions

• Results from this study suggest that metropolitan
  growth rate may be an important environmental
  factor related to obesity prevalence.
• We are currently extending this study to account
  for additional factors including transportation,
  built environment, climate, and culture.

2004 ISBNPA Conference June 10-13 Washington DC
Purpose

• This study compared the prevalence of obesity and
  leisure time physical activity (PA) in
  slowly-evolving cities versus rapidly-evolving
  cities.

Supported by the Centers for Disease Control and
Prevention Cooperative Agreement U48/CCU409664-06.